sample_id stringlengths 13 18 | original_caption stringlengths 20 250 | image_path stringlengths 55 55 | entities listlengths 1 16 | entity_foils listlengths 0 32 | relations_json stringlengths 2 3.58k | swap_negatives listlengths 1 4 |
|---|---|---|---|---|---|---|
coco_60060_314720 | a close up of a device on a toilet bowl | datasets/COCO/train2014/COCO_train2014_000000213069.jpg | [
"close up",
"device",
"toilet bowl"
] | [
{
"positive": "device",
"negative": "toilet paper holder",
"change_type": "object_change"
},
{
"positive": "device",
"negative": "toilet brush",
"change_type": "object_change"
},
{
"positive": "toilet bowl",
"negative": "toilet seat",
"change_type": "object_change"
},
... | [{"subject": "device", "relation_type": "is on", "object": "toilet bowl", "negatives": [{"subject": "device", "relation_type": "is under", "object": "toilet bowl", "change_type": "antonym"}, {"subject": "device", "relation_type": "is not on", "object": "toilet bowl", "change_type": "negation"}, {"subject": "toilet bowl... | [
"a close up of a toilet on a bowl device",
"a close up a on device of a toilet bowl",
"close a up device a of a on toilet bowl",
"bowl of a device on a toilet a close up"
] |
coco_60060_321503 | A ceramic of some sort with a logo of some sort. | datasets/COCO/train2014/COCO_train2014_000000213069.jpg | [
"ceramic of some sort",
"logo of some sort"
] | [
{
"positive": "logo of some sort",
"negative": "logo of a different sort",
"change_type": "attribute_change"
}
] | [{"subject": "ceramic of some sort", "relation_type": "has", "object": "logo of some sort", "negatives": [{"subject": "ceramic of some sort", "relation_type": "is without", "object": "logo of some sort", "change_type": "antonym"}, {"subject": "ceramic of some sort", "relation_type": "does not have", "object": "logo of ... | [
"A sort of some ceramic with a sort of some logo .",
"some ceramic with some sort a A logo of . sort of",
"A ceramic of sort some with of logo a sort some .",
"some sort . A ceramic of some sort with a logo of"
] |
coco_60060_322709 | A white tray on a toilet in bathroom on tiled floor. | datasets/COCO/train2014/COCO_train2014_000000213069.jpg | [
"white tray",
"toilet",
"bathroom",
"tiled floor"
] | [
{
"positive": "white tray",
"negative": "black tray",
"change_type": "attribute_change"
},
{
"positive": "white tray",
"negative": "white bowl",
"change_type": "object_change"
},
{
"positive": "toilet",
"negative": "sink",
"change_type": "object_change"
},
{
"posi... | [{"subject": "white tray", "relation_type": "is on", "object": "toilet", "negatives": [{"subject": "white tray", "relation_type": "is under", "object": "toilet", "change_type": "antonym"}, {"subject": "white tray", "relation_type": "is not on", "object": "toilet", "change_type": "negation"}, {"subject": "toilet", "rela... | [
"A white bathroom on a floor in tray on tiled toilet .",
"on white tray tiled . toilet in bathroom A a floor on",
"white A tray toilet on a bathroom on in tiled . floor",
"A white tray on a toilet in bathroom on tiled floor ."
] |
coco_60061_651933 | Modern commode displayed in small tiled bathroom stall. | datasets/COCO/train2014/COCO_train2014_000000324443.jpg | [
"Modern commode",
"small tiled bathroom stall"
] | [
{
"positive": "Modern commode",
"negative": "antique commode",
"change_type": "attribute_change"
},
{
"positive": "Modern commode",
"negative": "modern toilet",
"change_type": "object_change"
},
{
"positive": "small tiled bathroom stall",
"negative": "large wooden bathroom st... | [{"subject": "Modern commode", "relation_type": "is displayed in", "object": "small tiled bathroom stall", "negatives": [{"subject": "Modern commode", "relation_type": "is not displayed in", "object": "small tiled bathroom stall", "change_type": "negation"}, {"subject": "small tiled bathroom stall", "relation_type": "i... | [
"stall Modern displayed in small commode tiled bathroom .",
"Modern commode in . small tiled bathroom stall displayed",
"displayed commode Modern in small tiled bathroom . stall",
"Modern commode displayed in small tiled bathroom stall ."
] |
coco_60061_651960 | This toilet has a weird plastic piece on it. | datasets/COCO/train2014/COCO_train2014_000000324443.jpg | [
"toilet",
"weird plastic piece"
] | [
{
"positive": "toilet",
"negative": "sink",
"change_type": "object_change"
},
{
"positive": "weird plastic piece",
"negative": "metal bracket",
"change_type": "object_change"
},
{
"positive": "weird plastic piece",
"negative": "weird wooden piece",
"change_type": "attribu... | [{"subject": "toilet", "relation_type": "has", "object": "weird plastic piece", "negatives": [{"subject": "toilet", "relation_type": "is without", "object": "weird plastic piece", "change_type": "antonym"}, {"subject": "toilet", "relation_type": "does not have", "object": "weird plastic piece", "change_type": "negation... | [
"This piece has a plastic weird toilet on it .",
"This toilet has on weird plastic piece . it a",
"toilet has This plastic a weird piece on it .",
". piece on it This toilet has a weird plastic"
] |
coco_60061_654693 | A close up view looking down into a white toilet. | datasets/COCO/train2014/COCO_train2014_000000324443.jpg | [
"close up view",
"white toilet"
] | [
{
"positive": "white toilet",
"negative": "black toilet",
"change_type": "attribute_change"
},
{
"positive": "white toilet",
"negative": "white sink",
"change_type": "object_change"
}
] | [{"subject": "close up view", "relation_type": "is looking down into", "object": "white toilet", "negatives": [{"subject": "close up view", "relation_type": "is not looking down into", "object": "white toilet", "change_type": "negation"}, {"subject": "white toilet", "relation_type": "is looking down into", "object": "c... | [
"A close up white looking down into a view toilet .",
"into close . view A down up looking white toilet a",
"close up A view looking down white a into toilet .",
"into a white A close up toilet . view looking down"
] |
coco_60061_657264 | A toilet lid with a concentric design on it. | datasets/COCO/train2014/COCO_train2014_000000324443.jpg | [
"toilet lid",
"concentric design"
] | [
{
"positive": "toilet lid",
"negative": "toilet seat",
"change_type": "object_change"
},
{
"positive": "toilet lid",
"negative": "bathroom lid",
"change_type": "object_change"
},
{
"positive": "concentric design",
"negative": "geometric pattern",
"change_type": "attribute... | [{"subject": "toilet lid", "relation_type": "has", "object": "concentric design", "negatives": [{"subject": "toilet lid", "relation_type": "is without", "object": "concentric design", "change_type": "antonym"}, {"subject": "toilet lid", "relation_type": "does not have", "object": "concentric design", "change_type": "ne... | [
"A lid design with a toilet concentric on it .",
". toilet lid A on concentric design a with it",
"A lid toilet with concentric a it design on .",
"with a concentric design on it . A toilet lid"
] |
coco_60061_658074 | a close up of a device on a toilet in a bath room | datasets/COCO/train2014/COCO_train2014_000000324443.jpg | [
"close up",
"device",
"toilet",
"bath room"
] | [
{
"positive": "toilet",
"negative": "sink",
"change_type": "object_change"
},
{
"positive": "bath room",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "bath room",
"negative": "bathroom mirror",
"change_type": "attribute_change"
}
] | [{"subject": "close up", "relation_type": "is of", "object": "device", "negatives": []}, {"subject": "device", "relation_type": "is on", "object": "toilet", "negatives": []}, {"subject": "toilet", "relation_type": "is in", "object": "bath room", "negatives": []}] | [
"a close up of a room on a toilet in a bath device",
"of close a on a device a a toilet in up bath room",
"close up a device a of on a toilet in a bath room",
"on a toilet a close up of a device room in a bath"
] |
coco_60062_571410 | A white toilet sitting next to a parking meter. | datasets/COCO/train2014/COCO_train2014_000000356154.jpg | [
"white toilet",
"parking meter"
] | [
{
"positive": "white toilet",
"negative": "black toilet",
"change_type": "attribute_change"
},
{
"positive": "white toilet",
"negative": "white sink",
"change_type": "object_change"
},
{
"positive": "parking meter",
"negative": "fire hydrant",
"change_type": "object_chang... | [{"subject": "white toilet", "relation_type": "is sitting next to", "object": "parking meter", "negatives": [{"subject": "white toilet", "relation_type": "is away from", "object": "parking meter", "change_type": "antonym"}, {"subject": "white toilet", "relation_type": "is not sitting next to", "object": "parking meter"... | [
"A meter toilet sitting next to a white parking .",
"sitting white toilet a next A to parking meter .",
"white A toilet to next sitting parking meter a .",
"a parking meter . A white toilet sitting next to"
] |
coco_60062_572022 | a toilet next to a parking meter | datasets/COCO/train2014/COCO_train2014_000000356154.jpg | [
"toilet",
"parking meter"
] | [
{
"positive": "toilet",
"negative": "bench",
"change_type": "object_change"
}
] | [{"subject": "toilet", "relation_type": "is next to", "object": "parking meter", "negatives": [{"subject": "toilet", "relation_type": "is away from", "object": "parking meter", "change_type": "antonym"}, {"subject": "toilet", "relation_type": "is not next to", "object": "parking meter", "change_type": "negation"}, {"su... | [
"a meter next to a toilet parking",
"next toilet a a to parking meter",
"toilet next a a to parking meter",
"meter a toilet next to a parking"
] |
coco_60062_575937 | a toilet is out on the sidewalk next to a machine | datasets/COCO/train2014/COCO_train2014_000000356154.jpg | [
"toilet",
"sidewalk",
"machine"
] | [
{
"positive": "toilet",
"negative": "bench",
"change_type": "object_change"
},
{
"positive": "sidewalk",
"negative": "park",
"change_type": "object_change"
}
] | [{"subject": "toilet", "relation_type": "is out on", "object": "sidewalk", "negatives": [{"subject": "toilet", "relation_type": "is inside", "object": "sidewalk", "change_type": "antonym"}, {"subject": "toilet", "relation_type": "is not out on", "object": "sidewalk", "change_type": "negation"}]}, {"subject": "toilet", ... | [
"a machine is out on the toilet next to a sidewalk",
"to toilet on next the out sidewalk is a a machine",
"toilet is a the on out next to sidewalk machine a",
"sidewalk next to out on the a toilet is a machine"
] |
coco_60062_576006 | A toilet is sitting out in the public on the sidewalk. | datasets/COCO/train2014/COCO_train2014_000000356154.jpg | [
"toilet",
"public",
"sidewalk"
] | [
{
"positive": "toilet",
"negative": "bench",
"change_type": "object_change"
},
{
"positive": "sidewalk",
"negative": "park",
"change_type": "object_change"
},
{
"positive": "sidewalk",
"negative": "street",
"change_type": "object_change"
}
] | [{"subject": "toilet", "relation_type": "is sitting out in", "object": "public", "negatives": [{"subject": "toilet", "relation_type": "is not sitting out in", "object": "public", "change_type": "negation"}, {"subject": "toilet", "relation_type": "is inside", "object": "public", "change_type": "antonym"}]}, {"subject": ... | [
"A toilet is sitting out in the sidewalk on the public .",
". toilet on the out A sitting public the is sidewalk in",
"is A toilet sitting out in on the public . sidewalk the",
"sitting out in the sidewalk . A toilet is the public on"
] |
coco_60062_576099 | a white toilet next to a large black box | datasets/COCO/train2014/COCO_train2014_000000356154.jpg | [
"a white toilet",
"a large black box"
] | [
{
"positive": "a white toilet",
"negative": "a black toilet",
"change_type": "attribute_change"
},
{
"positive": "a white toilet",
"negative": "a white chair",
"change_type": "object_change"
},
{
"positive": "a large black box",
"negative": "a small black box",
"change_ty... | [{"subject": "a white toilet", "relation_type": "is next to", "object": "a large black box", "negatives": [{"subject": "a white toilet", "relation_type": "is away from", "object": "a large black box", "change_type": "antonym"}, {"subject": "a white toilet", "relation_type": "is not next to", "object": "a large black bo... | [
"a large white next to a box black toilet",
"toilet a white next to a large black box",
"next to a a white toilet large black box"
] |
coco_60063_165592 | A row of multicolored motorcycles parked next to a fence. | datasets/COCO/train2014/COCO_train2014_000000286886.jpg | [
"row of multicolored motorcycles",
"fence"
] | [
{
"positive": "row of multicolored motorcycles",
"negative": "row of black motorcycles",
"change_type": "attribute_change"
},
{
"positive": "row of multicolored motorcycles",
"negative": "row of multicolored bicycles",
"change_type": "object_change"
},
{
"positive": "fence",
... | [{"subject": "row of multicolored motorcycles", "relation_type": "is parked next to", "object": "fence", "negatives": [{"subject": "row of multicolored motorcycles", "relation_type": "is away from", "object": "fence", "change_type": "antonym"}, {"subject": "row of multicolored motorcycles", "relation_type": "is not par... | [
"A motorcycles of multicolored fence parked next to a row .",
"a row of to motorcycles multicolored . next A fence parked",
"row of A multicolored parked motorcycles a next to fence .",
"A row of fence . next to a multicolored motorcycles parked"
] |
coco_60063_167062 | A group of motorcycles parked next to each other in front of a fence | datasets/COCO/train2014/COCO_train2014_000000286886.jpg | [
"group of motorcycles",
"fence"
] | [
{
"positive": "group of motorcycles",
"negative": "group of bicycles",
"change_type": "object_change"
},
{
"positive": "fence",
"negative": "gate",
"change_type": "object_change"
}
] | [{"subject": "group of motorcycles", "relation_type": "are parked next to", "object": "fence", "negatives": [{"subject": "group of motorcycles", "relation_type": "are parked away from", "object": "fence", "change_type": "antonym"}, {"subject": "group of motorcycles", "relation_type": "are not parked next to", "object":... | [
"A other of front parked next to each group in motorcycles of a fence",
"parked group to motorcycles of in of next other A front each a fence",
"group of A motorcycles parked next other each to of front in a fence",
"to each other motorcycles parked next in front of A group of a fence"
] |
coco_60063_170386 | An assortment of motorcycles are parked near a fence. | datasets/COCO/train2014/COCO_train2014_000000286886.jpg | [
"an assortment of motorcycles",
"fence"
] | [
{
"positive": "an assortment of motorcycles",
"negative": "an assortment of bicycles",
"change_type": "object_change"
},
{
"positive": "an assortment of motorcycles",
"negative": "an assortment of trucks",
"change_type": "object_change"
},
{
"positive": "fence",
"negative": "... | [{"subject": "an assortment of motorcycles", "relation_type": "are parked near", "object": "fence", "negatives": [{"subject": "an assortment of motorcycles", "relation_type": "are far from", "object": "fence", "change_type": "antonym"}, {"subject": "an assortment of motorcycles", "relation_type": "are not parked near",... | [
"An fence of assortment are parked near a motorcycles .",
"are assortment of motorcycles a parked . near fence An",
"of assortment An are parked motorcycles near fence a .",
". motorcycles are parked near a fence An assortment of"
] |
coco_60063_173014 | Four motorcycles lined up in a parking area. | datasets/COCO/train2014/COCO_train2014_000000286886.jpg | [
"four motorcycles",
"parking area"
] | [
{
"positive": "four motorcycles",
"negative": "two motorcycles",
"change_type": "attribute_change"
},
{
"positive": "four motorcycles",
"negative": "four cars",
"change_type": "object_change"
}
] | [{"subject": "four motorcycles", "relation_type": "are lined up in", "object": "parking area", "negatives": [{"subject": "four motorcycles", "relation_type": "are scattered in", "object": "parking area", "change_type": "antonym"}, {"subject": "four motorcycles", "relation_type": "are not lined up in", "object": "parkin... | [
"Four area lined up in a parking motorcycles .",
". motorcycles in lined Four up parking area a",
"motorcycles lined Four a in up area parking .",
"parking area . Four motorcycles lined up in a"
] |
coco_60063_174181 | Motorcycles are parked in a row alongside a fenced in area. | datasets/COCO/train2014/COCO_train2014_000000286886.jpg | [
"motorcycles",
"fenced in area"
] | [
{
"positive": "motorcycles",
"negative": "cars",
"change_type": "object_change"
},
{
"positive": "fenced in area",
"negative": "open field",
"change_type": "object_change"
},
{
"positive": "fenced in area",
"negative": "fenced in garden",
"change_type": "attribute_change"... | [{"subject": "motorcycles", "relation_type": "are parked in a row alongside", "object": "fenced in area", "negatives": [{"subject": "motorcycles", "relation_type": "are not parked in a row alongside", "object": "fenced in area", "change_type": "negation"}, {"subject": "motorcycles", "relation_type": "are parked in a ro... | [
"Motorcycles are parked in a row alongside a fenced in area .",
"Motorcycles in fenced a parked row a are in . area alongside",
"parked Motorcycles are row a in a fenced alongside . in area",
"Motorcycles are parked in area . in a row alongside a fenced"
] |
coco_60064_372913 | Front of a Harley-Davidson store with motorcycles parked outside of it. | datasets/COCO/train2014/COCO_train2014_000000555118.jpg | [
"Front of a Harley-Davidson store",
"motorcycles"
] | [
{
"positive": "Front of a Harley-Davidson store",
"negative": "Front of a Honda motorcycle store",
"change_type": "object_change"
},
{
"positive": "motorcycles",
"negative": "cars",
"change_type": "object_change"
},
{
"positive": "motorcycles",
"negative": "bicycles",
"ch... | [{"subject": "motorcycles", "relation_type": "are parked outside of", "object": "Front of a Harley-Davidson store", "negatives": [{"subject": "motorcycles", "relation_type": "are inside of", "object": "Front of a Harley-Davidson store", "change_type": "antonym"}, {"subject": "motorcycles", "relation_type": "are not par... | [
"Front of a Harley - store motorcycles with Davidson parked outside of it .",
"Front a it Harley outside Davidson store of motorcycles of - . parked with",
"Front of a store Harley-Davidson with motorcycles outside parked . of it",
"Front of a Harley-Davidson store with motorcycles parked outside of it ."
] |
coco_60064_372946 | A group of people walking next to shops on a street. | datasets/COCO/train2014/COCO_train2014_000000555118.jpg | [
"group of people",
"shops",
"street"
] | [
{
"positive": "group of people",
"negative": "group of animals",
"change_type": "object_change"
},
{
"positive": "shops",
"negative": "cafés",
"change_type": "object_change"
},
{
"positive": "street",
"negative": "forest",
"change_type": "object_change"
},
{
"posi... | [{"subject": "group of people", "relation_type": "is walking next to", "object": "shops", "negatives": [{"subject": "group of people", "relation_type": "is walking away from", "object": "shops", "change_type": "antonym"}, {"subject": "group of people", "relation_type": "is not walking next to", "object": "shops", "chan... | [
"A shops of street walking next to people on a group .",
". group to people a A of shops next on street walking",
"A group of people walking next shops on to street a .",
"people walking next a street . to shops on A group of"
] |
coco_60064_373852 | Front of a Harley-Davidson store on a crowded retail street. | datasets/COCO/train2014/COCO_train2014_000000555118.jpg | [
"Front of a Harley-Davidson store",
"crowded retail street"
] | [
{
"positive": "Front of a Harley-Davidson store",
"negative": "Front of a Toyota dealership",
"change_type": "object_change"
},
{
"positive": "crowded retail street",
"negative": "empty retail street",
"change_type": "attribute_change"
},
{
"positive": "crowded retail street",
... | [{"subject": "Front of a Harley-Davidson store", "relation_type": "is on", "object": "crowded retail street", "negatives": [{"subject": "Front of a Harley-Davidson store", "relation_type": "is under", "object": "crowded retail street", "change_type": "antonym"}, {"subject": "Front of a Harley-Davidson store", "relation... | [
"Front of a Harley - crowded store on a Davidson retail street .",
"Front on a Harley of Davidson store - . crowded retail street a",
"of a Front on store Harley-Davidson crowded retail a street .",
"a crowded retail street . Front of a Harley-Davidson store on"
] |
coco_60064_376015 | Pedestrians walk past a motorcycle shop on a city street. | datasets/COCO/train2014/COCO_train2014_000000555118.jpg | [
"pedestrians",
"motorcycle shop",
"city street"
] | [
{
"positive": "pedestrians",
"negative": "cyclists",
"change_type": "object_change"
},
{
"positive": "motorcycle shop",
"negative": "car shop",
"change_type": "object_change"
},
{
"positive": "motorcycle shop",
"negative": "motorcycle repair",
"change_type": "attribute_ch... | [{"subject": "pedestrians", "relation_type": "are walking past", "object": "motorcycle shop", "negatives": [{"subject": "pedestrians", "relation_type": "are walking away from", "object": "motorcycle shop", "change_type": "antonym"}, {"subject": "pedestrians", "relation_type": "are not walking past", "object": "motorcyc... | [
"shop walk past a city street on a motorcycle Pedestrians .",
"Pedestrians a on past motorcycle shop . walk city street a",
"past walk Pedestrians shop motorcycle a city on a . street",
"a motorcycle shop street . Pedestrians walk past on a city"
] |
coco_60064_377497 | The front of a Harley Davidson store in Paris. | datasets/COCO/train2014/COCO_train2014_000000555118.jpg | [
"front of a Harley Davidson store",
"Paris"
] | [
{
"positive": "front of a Harley Davidson store",
"negative": "front of a Renault store",
"change_type": "object_change"
},
{
"positive": "Paris",
"negative": "Rome",
"change_type": "object_change"
},
{
"positive": "Paris",
"negative": "London",
"change_type": "object_cha... | [{"subject": "front of a Harley Davidson store", "relation_type": "is in", "object": "Paris", "negatives": [{"subject": "front of a Harley Davidson store", "relation_type": "is not in", "object": "Paris", "change_type": "negation"}, {"subject": "front of a Harley Davidson store", "relation_type": "is outside", "object"... | [
"The store of a Paris front Davidson in Harley .",
"The front . a Harley Davidson store in Paris of",
"of front The a Harley Davidson Paris in store .",
"The front of a Harley Davidson store in Paris ."
] |
coco_60065_147752 | A man and woman in a restroom with the toilet spraying water. | datasets/COCO/train2014/COCO_train2014_000000197243.jpg | [
"man",
"woman",
"restroom",
"toilet",
"water"
] | [
{
"positive": "restroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "toilet",
"negative": "sink",
"change_type": "object_change"
},
{
"positive": "toilet",
"negative": "shower",
"change_type": "object_change"
},
{
"positive": "water",
... | [{"subject": "man", "relation_type": "is in", "object": "restroom", "negatives": []}, {"subject": "woman", "relation_type": "is in", "object": "restroom", "negatives": []}, {"subject": "toilet", "relation_type": "is in", "object": "restroom", "negatives": []}, {"subject": "toilet", "relation_type": "is spraying", "obje... | [
"A woman and man in a water with the restroom spraying toilet .",
"in man spraying woman with a restroom and . toilet the water A",
"and man A a in woman with restroom the water spraying toilet .",
"toilet spraying water woman in a A man and restroom with the ."
] |
coco_60065_149162 | Two people pose in a bathroom while water squirts from a toilet. | datasets/COCO/train2014/COCO_train2014_000000197243.jpg | [
"two people",
"bathroom",
"water",
"toilet"
] | [
{
"positive": "two people",
"negative": "three people",
"change_type": "attribute_change"
},
{
"positive": "two people",
"negative": "two dogs",
"change_type": "object_change"
},
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
... | [{"subject": "two people", "relation_type": "are posing in", "object": "bathroom", "negatives": [{"subject": "two people", "relation_type": "are not posing in", "object": "bathroom", "change_type": "negation"}, {"subject": "bathroom", "relation_type": "are posing in", "object": "two people", "change_type": "subject_obj... | [
"Two people pose in a water while squirts toilet from a bathroom .",
"in people a a . bathroom Two water squirts while from toilet pose",
"Two people pose in bathroom a water squirts while a from toilet .",
"in a bathroom . Two people pose while water squirts from a toilet"
] |
coco_60065_154685 | A pretty blonde hugging a man in a bathroom. | datasets/COCO/train2014/COCO_train2014_000000197243.jpg | [
"pretty blonde",
"man",
"bathroom"
] | [
{
"positive": "pretty blonde",
"negative": "tall brunette",
"change_type": "attribute_change"
},
{
"positive": "pretty blonde",
"negative": "pretty brunette",
"change_type": "attribute_change"
},
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_chan... | [{"subject": "pretty blonde", "relation_type": "is hugging", "object": "man", "negatives": [{"subject": "pretty blonde", "relation_type": "is pushing away", "object": "man", "change_type": "antonym"}, {"subject": "pretty blonde", "relation_type": "is not hugging", "object": "man", "change_type": "negation"}, {"subject"... | [
"A pretty man hugging a blonde in a bathroom .",
"hugging in blonde a pretty man A . bathroom a",
"pretty A blonde hugging a man bathroom a in .",
"A pretty blonde in a bathroom hugging a man ."
] |
coco_60065_158843 | two people in a bath room that seems to be leaking | datasets/COCO/train2014/COCO_train2014_000000197243.jpg | [
"two people",
"bathroom",
"leaking"
] | [
{
"positive": "two people",
"negative": "three people",
"change_type": "attribute_change"
},
{
"positive": "two people",
"negative": "two dogs",
"change_type": "object_change"
},
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
... | [{"subject": "bathroom", "relation_type": "is leaking", "object": "leaking", "negatives": [{"subject": "bathroom", "relation_type": "is not leaking", "object": "leaking", "change_type": "negation"}]}, {"subject": "two people", "relation_type": "are in", "object": "bathroom", "negatives": [{"subject": "two people", "rel... | [
"two room in a people bath that seems to be leaking",
"in people a seems bath room that leaking two be to",
"people two in a room bath to that seems leaking be",
"a bath room that seems to two people in be leaking"
] |
coco_60065_159233 | Man and woman posing in a bathroom with a broken toilet. | datasets/COCO/train2014/COCO_train2014_000000197243.jpg | [
"man",
"woman",
"bathroom",
"broken toilet"
] | [
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "broken toilet",
"negative": "working toilet",
"change_type": "attribute_change"
},
{
"positive": "broken toilet",
"negative": "broken sink",
"change_type": "object_change"
... | [{"subject": "man", "relation_type": "is posing in", "object": "bathroom", "negatives": [{"subject": "man", "relation_type": "is not posing in", "object": "bathroom", "change_type": "negation"}]}, {"subject": "woman", "relation_type": "is posing in", "object": "bathroom", "negatives": [{"subject": "woman", "relation_ty... | [
"woman and broken posing in a bathroom with a Man toilet .",
"Man in woman a posing . bathroom and with broken toilet a",
"and Man woman posing a in a with bathroom broken toilet .",
"broken toilet . bathroom with a Man and woman posing in a"
] |
coco_60066_334684 | there is a all white motorcycle on display | datasets/COCO/train2014/COCO_train2014_000000077569.jpg | [
"all white motorcycle"
] | [
{
"positive": "all white motorcycle",
"negative": "all black motorcycle",
"change_type": "attribute_change"
},
{
"positive": "all white motorcycle",
"negative": "all white car",
"change_type": "object_change"
}
] | [] | [
"there is a all motorcycle white on display",
"on is all there white motorcycle a display",
"there a is all motorcycle white on display",
"on display all white motorcycle there is a"
] |
coco_60066_334720 | A close up of a modern motorcycle on display. | datasets/COCO/train2014/COCO_train2014_000000077569.jpg | [
"close up",
"modern motorcycle",
"display"
] | [
{
"positive": "modern motorcycle",
"negative": "classic motorcycle",
"change_type": "attribute_change"
},
{
"positive": "modern motorcycle",
"negative": "modern car",
"change_type": "object_change"
}
] | [{"subject": "close up", "relation_type": "is of", "object": "modern motorcycle", "negatives": [{"subject": "close up", "relation_type": "is not of", "object": "modern motorcycle", "change_type": "negation"}]}, {"subject": "modern motorcycle", "relation_type": "is on", "object": "display", "negatives": [{"subject": "mo... | [
"A modern up of a motorcycle close on display .",
"of close up on a modern motorcycle A display .",
"A up close modern a of display motorcycle on .",
"A close up motorcycle on display of a modern ."
] |
coco_60066_335413 | a white motorcycle sittting in front of a white sign | datasets/COCO/train2014/COCO_train2014_000000077569.jpg | [
"white motorcycle",
"white sign"
] | [
{
"positive": "white motorcycle",
"negative": "black motorcycle",
"change_type": "attribute_change"
},
{
"positive": "white motorcycle",
"negative": "white car",
"change_type": "object_change"
},
{
"positive": "white sign",
"negative": "black sign",
"change_type": "attrib... | [{"subject": "white motorcycle", "relation_type": "is sitting in front of", "object": "white sign", "negatives": [{"subject": "white motorcycle", "relation_type": "is sitting behind", "object": "white sign", "change_type": "antonym"}, {"subject": "white motorcycle", "relation_type": "is not sitting in front of", "objec... | [
"a white sign sittting in motorcycle of a white front",
"of white motorcycle a sittting front a in white sign",
"a white motorcycle front in sittting a of white sign",
"sign of a white a white motorcycle sittting in front"
] |
coco_60066_339631 | A large white motor bike on display parked. | datasets/COCO/train2014/COCO_train2014_000000077569.jpg | [
"large white motor bike",
"display"
] | [
{
"positive": "large white motor bike",
"negative": "small black motor bike",
"change_type": "attribute_change"
},
{
"positive": "large white motor bike",
"negative": "large white car",
"change_type": "object_change"
}
] | [{"subject": "large white motor bike", "relation_type": "is on display", "object": "display", "negatives": []}] | [
"A motor display white bike on large parked .",
"parked large white motor bike on display . A",
"A large white motor bike on parked . display",
"A large white motor bike on display parked ."
] |
coco_60066_342892 | A white concept motorcycle on a black display. | datasets/COCO/train2014/COCO_train2014_000000077569.jpg | [
"white concept motorcycle",
"black display"
] | [
{
"positive": "white concept motorcycle",
"negative": "black concept motorcycle",
"change_type": "attribute_change"
},
{
"positive": "white concept motorcycle",
"negative": "white concept car",
"change_type": "object_change"
},
{
"positive": "black display",
"negative": "whit... | [{"subject": "white concept motorcycle", "relation_type": "is on", "object": "black display", "negatives": [{"subject": "white concept motorcycle", "relation_type": "is under", "object": "black display", "change_type": "antonym"}, {"subject": "white concept motorcycle", "relation_type": "is not on", "object": "black di... | [
"A display black white on a motorcycle concept .",
"A white concept motorcycle on . black display a",
"concept white A motorcycle a on black . display",
"motorcycle on a black display . A white concept"
] |
coco_60067_325345 | Some double decker buses parked next to each other by a building. | datasets/COCO/train2014/COCO_train2014_000000286190.jpg | [
"double decker buses",
"building"
] | [
{
"positive": "double decker buses",
"negative": "single decker buses",
"change_type": "attribute_change"
},
{
"positive": "double decker buses",
"negative": "double decker trucks",
"change_type": "object_change"
},
{
"positive": "building",
"negative": "tree",
"change_ty... | [{"subject": "double decker buses", "relation_type": "are parked next to", "object": "building", "negatives": [{"subject": "double decker buses", "relation_type": "are parked away from", "object": "building", "change_type": "antonym"}, {"subject": "double decker buses", "relation_type": "are not parked next to", "objec... | [
"Some double decker building parked next to each buses by a other .",
"Some double decker buses parked a each to other . by building next",
"double Some decker parked buses next each to other by building a .",
"by a building buses parked next to each other . Some double decker"
] |
coco_60067_326644 | three red double stacker buses parked behind one another | datasets/COCO/train2014/COCO_train2014_000000286190.jpg | [
"three red double stacker buses",
"one another"
] | [
{
"positive": "three red double stacker buses",
"negative": "two blue double stacker buses",
"change_type": "attribute_change"
},
{
"positive": "three red double stacker buses",
"negative": "three red garbage trucks",
"change_type": "object_change"
}
] | [{"subject": "three red double stacker buses", "relation_type": "are parked behind", "object": "one another", "negatives": [{"subject": "three red double stacker buses", "relation_type": "are parked in front of", "object": "one another", "change_type": "antonym"}, {"subject": "three red double stacker buses", "relation... | [
"three double stacker red buses parked behind one another",
" red double stacker buses one behind three parked another",
"double three red parked stacker buses one behind another",
"stacker buses parked behind one another three red double"
] |
coco_60067_329689 | A line of double decker buses parked along a roadway. | datasets/COCO/train2014/COCO_train2014_000000286190.jpg | [
"line of double decker buses",
"roadway"
] | [
{
"positive": "line of double decker buses",
"negative": "row of fire trucks",
"change_type": "object_change"
},
{
"positive": "line of double decker buses",
"negative": "line of single decker buses",
"change_type": "attribute_change"
},
{
"positive": "roadway",
"negative": "... | [{"subject": "line of double decker buses", "relation_type": "is parked along", "object": "roadway", "negatives": [{"subject": "line of double decker buses", "relation_type": "is not parked along", "object": "roadway", "change_type": "negation"}, {"subject": "roadway", "relation_type": "is parked along", "object": "lin... | [
"A double of buses line decker parked along a roadway .",
"parked line a double decker buses along of A roadway .",
"line A of double decker buses along parked a roadway .",
"A line of double decker buses roadway . parked along a"
] |
coco_60067_330700 | A line of red, double decked buses line up against a sidewalk near old, brown, noble buildings. | datasets/COCO/train2014/COCO_train2014_000000286190.jpg | [
"a line of red, double decked buses",
"sidewalk",
"old, brown, noble buildings"
] | [
{
"positive": "sidewalk",
"negative": "parking lot",
"change_type": "object_change"
},
{
"positive": "old, brown, noble buildings",
"negative": "new, white, modern buildings",
"change_type": "attribute_change"
},
{
"positive": "old, brown, noble buildings",
"negative": "old, ... | [{"subject": "a line of red, double decked buses", "relation_type": "is lining up against", "object": "sidewalk", "negatives": [{"subject": "a line of red, double decked buses", "relation_type": "is moving away from", "object": "sidewalk", "change_type": "antonym"}, {"subject": "a line of red, double decked buses", "re... | [
"A brown of buildings , buses old sidewalk line up against a noble near double , line , decked red .",
"line line A red against double decked buses near up . , sidewalk of old a brown , noble buildings ,",
"line A of double , red line buses decked up a against near sidewalk old , , brown noble buildings .",
"... |
coco_60067_335485 | there are many red double decker busses on this street | datasets/COCO/train2014/COCO_train2014_000000286190.jpg | [
"many red double decker busses",
"this street"
] | [
{
"positive": "many red double decker busses",
"negative": "many blue double decker busses",
"change_type": "attribute_change"
},
{
"positive": "many red double decker busses",
"negative": "many red trucks",
"change_type": "object_change"
},
{
"positive": "this street",
"nega... | [{"subject": "many red double decker busses", "relation_type": "are on", "object": "this street", "negatives": [{"subject": "many red double decker busses", "relation_type": "are under", "object": "this street", "change_type": "antonym"}, {"subject": "many red double decker busses", "relation_type": "are not on", "obje... | [
"there are decker street red many double on this busses",
"on are many red double decker busses there this street",
"many are there red decker double this busses on street",
"street red double decker busses on this there are many"
] |
coco_60068_285622 | A white motorcycle parked in a museum on a black floor. | datasets/COCO/train2014/COCO_train2014_000000053696.jpg | [
"white motorcycle",
"museum",
"black floor"
] | [
{
"positive": "white motorcycle",
"negative": "black motorcycle",
"change_type": "attribute_change"
},
{
"positive": "white motorcycle",
"negative": "white car",
"change_type": "object_change"
},
{
"positive": "museum",
"negative": "forest",
"change_type": "object_change"... | [{"subject": "white motorcycle", "relation_type": "is parked in", "object": "museum", "negatives": [{"subject": "white motorcycle", "relation_type": "is not parked in", "object": "museum", "change_type": "negation"}, {"subject": "museum", "relation_type": "is parked in", "object": "white motorcycle", "change_type": "su... | [
"A museum motorcycle parked in a black on a floor white .",
"on white motorcycle . a a museum parked in black floor A",
"motorcycle white A parked a in on a museum black floor .",
"black floor . parked in a A white motorcycle museum on a"
] |
coco_60068_401941 | A close up of a white concept motorcycle on display. | datasets/COCO/train2014/COCO_train2014_000000053696.jpg | [
"close up",
"white concept motorcycle",
"display"
] | [
{
"positive": "white concept motorcycle",
"negative": "black concept car",
"change_type": "object_change"
},
{
"positive": "white concept motorcycle",
"negative": "white concept bicycle",
"change_type": "object_change"
}
] | [{"subject": "close up", "relation_type": "is of", "object": "white concept motorcycle", "negatives": [{"subject": "close up", "relation_type": "is not of", "object": "white concept motorcycle", "change_type": "negation"}]}, {"subject": "white concept motorcycle", "relation_type": "is on", "object": "display", "negativ... | [
"A white up of a concept close motorcycle on display .",
"up close . A on white concept motorcycle a display of",
"up A close a of white on motorcycle concept display .",
"concept motorcycle on display . A close up of a white"
] |
coco_60068_405265 | A back view of a white and blue motorcycle. | datasets/COCO/train2014/COCO_train2014_000000053696.jpg | [
"back view of a white and blue motorcycle"
] | [
{
"positive": "back view of a white and blue motorcycle",
"negative": "back view of a red and black car",
"change_type": "object_change"
}
] | [] | [
"A view back of a blue and white motorcycle .",
"a back view and A white . blue motorcycle of",
"back view A a white of blue motorcycle and .",
". of a white A back view and blue motorcycle"
] |
coco_60068_406162 | Very aesthetic motorcycle on display with good lighting. | datasets/COCO/train2014/COCO_train2014_000000053696.jpg | [
"very aesthetic motorcycle",
"good lighting"
] | [
{
"positive": "very aesthetic motorcycle",
"negative": "very aesthetic bicycle",
"change_type": "object_change"
},
{
"positive": "very aesthetic motorcycle",
"negative": "very aesthetic car",
"change_type": "object_change"
},
{
"positive": "good lighting",
"negative": "poor l... | [{"subject": "very aesthetic motorcycle", "relation_type": "is on display with", "object": "good lighting", "negatives": [{"subject": "very aesthetic motorcycle", "relation_type": "is not on display with", "object": "good lighting", "change_type": "negation"}, {"subject": "very aesthetic motorcycle", "relation_type": "... | [
"Very display motorcycle on aesthetic with lighting good .",
". aesthetic motorcycle on display with good lighting Very",
"aesthetic motorcycle Very display on with good . lighting",
"on display with good lighting . Very aesthetic motorcycle"
] |
coco_60068_406306 | a motorcycle on display in front of a company sign | datasets/COCO/train2014/COCO_train2014_000000053696.jpg | [
"motorcycle",
"company sign"
] | [
{
"positive": "motorcycle",
"negative": "truck",
"change_type": "object_change"
},
{
"positive": "company sign",
"negative": "company logo",
"change_type": "object_change"
}
] | [{"subject": "motorcycle", "relation_type": "is on display in front of", "object": "company sign", "negatives": [{"subject": "motorcycle", "relation_type": "is on display behind", "object": "company sign", "change_type": "antonym"}, {"subject": "motorcycle", "relation_type": "is not on display in front of", "object": "... | [
"a sign on company in display of a motorcycle front",
"a motorcycle of display a front on in company sign",
"motorcycle a on display in front of company a sign",
"a motorcycle on of a company display in front sign"
] |
coco_60069_454587 | A teenaged boy lounges on a ledge in a small white bathroom. | datasets/COCO/train2014/COCO_train2014_000000157554.jpg | [
"teenaged boy",
"ledge",
"small white bathroom"
] | [
{
"positive": "teenaged boy",
"negative": "teenaged girl",
"change_type": "attribute_change"
},
{
"positive": "teenaged boy",
"negative": "adult man",
"change_type": "attribute_change"
},
{
"positive": "ledge",
"negative": "bench",
"change_type": "object_change"
},
{
... | [{"subject": "teenaged boy", "relation_type": "is lounging on", "object": "ledge", "negatives": [{"subject": "teenaged boy", "relation_type": "is standing on", "object": "ledge", "change_type": "antonym"}, {"subject": "teenaged boy", "relation_type": "is not lounging on", "object": "ledge", "change_type": "negation"}, ... | [
"A lounges ledge small on a teenaged in a bathroom white boy .",
"in teenaged boy lounges a a ledge . on small white bathroom A",
"teenaged A boy a on lounges a ledge in small white bathroom .",
"small white bathroom ledge in a . A teenaged boy lounges on a"
] |
coco_60069_456627 | A young man laying next to a white toilet. | datasets/COCO/train2014/COCO_train2014_000000157554.jpg | [
"young man",
"white toilet"
] | [
{
"positive": "young man",
"negative": "old man",
"change_type": "attribute_change"
},
{
"positive": "young man",
"negative": "young woman",
"change_type": "attribute_change"
},
{
"positive": "white toilet",
"negative": "black toilet",
"change_type": "attribute_change"
... | [{"subject": "young man", "relation_type": "is laying next to", "object": "white toilet", "negatives": [{"subject": "young man", "relation_type": "is away from", "object": "white toilet", "change_type": "antonym"}, {"subject": "young man", "relation_type": "is not laying next to", "object": "white toilet", "change_type... | [
"A man young laying next to a toilet white .",
". young man a next laying A white toilet to",
"man young A to next laying toilet white a .",
"A young man a white toilet laying next to ."
] |
coco_60069_457200 | A person laying in a room with a toilet. | datasets/COCO/train2014/COCO_train2014_000000157554.jpg | [
"person",
"room",
"toilet"
] | [
{
"positive": "room",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "toilet",
"negative": "bed",
"change_type": "object_change"
}
] | [{"subject": "person", "relation_type": "is laying in", "object": "room", "negatives": [{"subject": "person", "relation_type": "is standing in", "object": "room", "change_type": "antonym"}, {"subject": "person", "relation_type": "is not laying in", "object": "room", "change_type": "negation"}, {"subject": "room", "rela... | [
"A toilet laying in a person with a room .",
"a person in a A room . with toilet laying",
"person laying A in a room a toilet with .",
"in a room with a toilet . A person laying"
] |
coco_60069_457575 | A man is laying right next to the toilet. | datasets/COCO/train2014/COCO_train2014_000000157554.jpg | [
"man",
"toilet"
] | [
{
"positive": "toilet",
"negative": "bed",
"change_type": "object_change"
}
] | [{"subject": "man", "relation_type": "is laying next to", "object": "toilet", "negatives": [{"subject": "man", "relation_type": "is away from", "object": "toilet", "change_type": "antonym"}, {"subject": "man", "relation_type": "is not laying next to", "object": "toilet", "change_type": "negation"}, {"subject": "toilet"... | [
"A toilet is laying right next to the man .",
"to man . is right the A laying toilet next",
"A is man laying right next to toilet the .",
"to the toilet laying right next . A man is"
] |
coco_60069_822078 | A man lounges in a bathroom next to the toilet. | datasets/COCO/train2014/COCO_train2014_000000157554.jpg | [
"man",
"bathroom",
"toilet"
] | [
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "bathroom",
"negative": "living room",
"change_type": "object_change"
},
{
"positive": "toilet",
"negative": "bed",
"change_type": "object_change"
}
] | [{"subject": "man", "relation_type": "is in", "object": "bathroom", "negatives": [{"subject": "man", "relation_type": "is outside", "object": "bathroom", "change_type": "antonym"}, {"subject": "man", "relation_type": "is not in", "object": "bathroom", "change_type": "negation"}]}, {"subject": "man", "relation_type": "i... | [
"A bathroom lounges in a man next to the toilet .",
"the man to A . bathroom next in lounges toilet a",
"man A lounges a in bathroom next the to toilet .",
"A man lounges in a bathroom next to the toilet ."
] |
coco_60070_296425 | A black motorcycle parked in a road covered by leaves. | datasets/COCO/train2014/COCO_train2014_000000113455.jpg | [
"black motorcycle",
"road",
"leaves"
] | [
{
"positive": "black motorcycle",
"negative": "red motorcycle",
"change_type": "attribute_change"
},
{
"positive": "black motorcycle",
"negative": "black bicycle",
"change_type": "object_change"
},
{
"positive": "road",
"negative": "parking lot",
"change_type": "object_ch... | [{"subject": "black motorcycle", "relation_type": "is parked on", "object": "road", "negatives": [{"subject": "black motorcycle", "relation_type": "is not parked on", "object": "road", "change_type": "negation"}, {"subject": "black motorcycle", "relation_type": "is moving away from", "object": "road", "change_type": "a... | [
"A road motorcycle parked in a leaves covered by black .",
"parked black motorcycle a . A road covered in leaves by",
"motorcycle black A parked a in covered road by . leaves",
"road covered by A black motorcycle leaves . parked in a"
] |
coco_60070_298021 | Motorcycle sitting on leaves in a parking lot. | datasets/COCO/train2014/COCO_train2014_000000113455.jpg | [
"motorcycle",
"leaves",
"parking lot"
] | [
{
"positive": "motorcycle",
"negative": "car",
"change_type": "object_change"
},
{
"positive": "leaves",
"negative": "rocks",
"change_type": "object_change"
},
{
"positive": "parking lot",
"negative": "parking garage",
"change_type": "object_change"
}
] | [{"subject": "motorcycle", "relation_type": "is sitting on", "object": "leaves", "negatives": [{"subject": "motorcycle", "relation_type": "is standing on", "object": "leaves", "change_type": "antonym"}, {"subject": "motorcycle", "relation_type": "is not sitting on", "object": "leaves", "change_type": "negation"}, {"sub... | [
"Motorcycle sitting on lot in a parking leaves .",
"Motorcycle a . leaves in on parking lot sitting",
"on Motorcycle sitting a leaves in lot parking .",
"parking lot . leaves in a Motorcycle sitting on"
] |
coco_60070_298792 | A motorcycle and several cars sitting next to a building. | datasets/COCO/train2014/COCO_train2014_000000113455.jpg | [
"motorcycle",
"several cars",
"building"
] | [
{
"positive": "several cars",
"negative": "multiple buses",
"change_type": "object_change"
},
{
"positive": "building",
"negative": "tree",
"change_type": "object_change"
},
{
"positive": "building",
"negative": "bridge",
"change_type": "object_change"
}
] | [{"subject": "motorcycle", "relation_type": "is sitting next to", "object": "building", "negatives": [{"subject": "motorcycle", "relation_type": "is away from", "object": "building", "change_type": "antonym"}, {"subject": "motorcycle", "relation_type": "is not sitting next to", "object": "building", "change_type": "neg... | [
"A building and motorcycle several sitting next to a cars .",
"a motorcycle A several cars sitting and to next building .",
"and motorcycle A sitting cars several a next to . building",
"next to a building . several cars sitting A motorcycle and"
] |
coco_60070_301003 | The motorcycle is parked in a pile of golden leaves. | datasets/COCO/train2014/COCO_train2014_000000113455.jpg | [
"motorcycle",
"pile of golden leaves"
] | [
{
"positive": "motorcycle",
"negative": "bicycle",
"change_type": "object_change"
},
{
"positive": "pile of golden leaves",
"negative": "pile of red leaves",
"change_type": "attribute_change"
},
{
"positive": "pile of golden leaves",
"negative": "pile of golden bricks",
"... | [{"subject": "motorcycle", "relation_type": "is parked in", "object": "pile of golden leaves", "negatives": [{"subject": "motorcycle", "relation_type": "is outside of", "object": "pile of golden leaves", "change_type": "antonym"}, {"subject": "motorcycle", "relation_type": "is not parked in", "object": "pile of golden ... | [
"The motorcycle is parked in a pile of leaves golden .",
"is motorcycle parked . The of pile a golden leaves in",
"motorcycle is The in a parked golden pile of leaves .",
"The motorcycle is pile of golden leaves . parked in a"
] |
coco_60070_304966 | A motorcycle that has been parked on some leaves. | datasets/COCO/train2014/COCO_train2014_000000113455.jpg | [
"motorcycle",
"leaves"
] | [
{
"positive": "motorcycle",
"negative": "car",
"change_type": "object_change"
},
{
"positive": "leaves",
"negative": "rocks",
"change_type": "object_change"
}
] | [{"subject": "motorcycle", "relation_type": "is parked on", "object": "leaves", "negatives": [{"subject": "motorcycle", "relation_type": "is under", "object": "leaves", "change_type": "antonym"}, {"subject": "motorcycle", "relation_type": "is not parked on", "object": "leaves", "change_type": "negation"}, {"subject": "... | [
"A leaves that has been parked on some motorcycle .",
"has motorcycle some on that . parked A leaves been",
"that motorcycle A has been parked on leaves some .",
"on some leaves has been parked A motorcycle that ."
] |
coco_60071_116362 | A picture of a very very dirty bathroom. | datasets/COCO/train2014/COCO_train2014_000000363194.jpg | [
"picture",
"very very dirty bathroom"
] | [
{
"positive": "very very dirty bathroom",
"negative": "very very clean bathroom",
"change_type": "attribute_change"
},
{
"positive": "very very dirty bathroom",
"negative": "very very dirty kitchen",
"change_type": "object_change"
}
] | [{"subject": "picture", "relation_type": "is of", "object": "very very dirty bathroom", "negatives": [{"subject": "picture", "relation_type": "is not of", "object": "very very dirty bathroom", "change_type": "negation"}]}] | [
"A picture of a very very dirty bathroom .",
"very picture very . of a dirty bathroom A",
"picture A of very very a . dirty bathroom",
"a very very dirty bathroom . A picture of"
] |
coco_60071_118555 | A destroyed toilet in a bathroom spraying water. | datasets/COCO/train2014/COCO_train2014_000000363194.jpg | [
"destroyed toilet",
"bathroom",
"water"
] | [
{
"positive": "destroyed toilet",
"negative": "intact toilet",
"change_type": "attribute_change"
},
{
"positive": "destroyed toilet",
"negative": "destroyed sink",
"change_type": "object_change"
},
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_ch... | [{"subject": "destroyed toilet", "relation_type": "is in", "object": "bathroom", "negatives": [{"subject": "destroyed toilet", "relation_type": "is not in", "object": "bathroom", "change_type": "negation"}, {"subject": "bathroom", "relation_type": "is in", "object": "destroyed toilet", "change_type": "subject_object_sw... | [
"A destroyed toilet in a bathroom spraying water .",
"a in toilet destroyed spraying bathroom A water .",
"toilet destroyed A bathroom a in water spraying .",
"spraying water . in a bathroom A destroyed toilet"
] |
coco_60071_119308 | Water leaking from a pipe in a vandalized public bathroom. | datasets/COCO/train2014/COCO_train2014_000000363194.jpg | [
"water",
"pipe",
"vandalized public bathroom"
] | [
{
"positive": "water",
"negative": "mud",
"change_type": "object_change"
},
{
"positive": "pipe",
"negative": "leaking faucet",
"change_type": "object_change"
},
{
"positive": "vandalized public bathroom",
"negative": "clean public bathroom",
"change_type": "attribute_cha... | [{"subject": "water", "relation_type": "is leaking from", "object": "pipe", "negatives": [{"subject": "water", "relation_type": "is not leaking from", "object": "pipe", "change_type": "negation"}, {"subject": "water", "relation_type": "is flowing into", "object": "pipe", "change_type": "antonym"}]}, {"subject": "pipe",... | [
"Water leaking from a public in a vandalized bathroom pipe .",
"Water vandalized in leaking pipe . from a public bathroom a",
"leaking from Water in a pipe public a vandalized . bathroom",
"a pipe in Water leaking from a vandalized public bathroom ."
] |
coco_60071_131917 | A bathroom with graffiti on the walls and a broken toilet that is leaking water from the tank. | datasets/COCO/train2014/COCO_train2014_000000363194.jpg | [
"bathroom",
"graffiti",
"walls",
"broken toilet",
"water",
"tank"
] | [
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "graffiti",
"negative": "painting",
"change_type": "object_change"
},
{
"positive": "walls",
"negative": "ceilings",
"change_type": "object_change"
},
{
"positive": "w... | [{"subject": "graffiti", "relation_type": "is on", "object": "walls", "negatives": [{"subject": "graffiti", "relation_type": "is under", "object": "walls", "change_type": "antonym"}, {"subject": "graffiti", "relation_type": "is not on", "object": "walls", "change_type": "negation"}]}, {"subject": "broken toilet", "rela... | [
"A bathroom with broken on the water and a toilet tank that is leaking walls from the graffiti .",
"on bathroom and graffiti the the walls from leaking broken toilet is a . water A with tank that",
"bathroom with A on the graffiti walls a and broken that toilet is leaking water from the tank .",
"from the tan... |
coco_60071_133306 | The toilet in the bathroom has been broke. | datasets/COCO/train2014/COCO_train2014_000000363194.jpg | [
"toilet",
"bathroom"
] | [
{
"positive": "toilet",
"negative": "sink",
"change_type": "object_change"
},
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "bathroom",
"negative": "living room",
"change_type": "object_change"
}
] | [{"subject": "toilet", "relation_type": "is in", "object": "bathroom", "negatives": [{"subject": "toilet", "relation_type": "is not in", "object": "bathroom", "change_type": "negation"}, {"subject": "toilet", "relation_type": "is outside", "object": "bathroom", "change_type": "antonym"}, {"subject": "bathroom", "relati... | [
"The bathroom in the toilet has been broke .",
"in toilet been has bathroom . broke the The",
"in The toilet bathroom has the broke . been",
"the bathroom has The toilet in been broke ."
] |
coco_60072_347199 | A corner of a bathroom has a toilet and a telephone in it. | datasets/COCO/train2014/COCO_train2014_000000290361.jpg | [
"corner of a bathroom",
"toilet",
"telephone"
] | [
{
"positive": "corner of a bathroom",
"negative": "corner of a kitchen",
"change_type": "object_change"
},
{
"positive": "corner of a bathroom",
"negative": "corner of a bedroom",
"change_type": "object_change"
},
{
"positive": "toilet",
"negative": "sink",
"change_type":... | [{"subject": "corner of a bathroom", "relation_type": "has", "object": "toilet", "negatives": [{"subject": "corner of a bathroom", "relation_type": "does not have", "object": "toilet", "change_type": "negation"}, {"subject": "toilet", "relation_type": "has", "object": "corner of a bathroom", "change_type": "subject_obj... | [
"A telephone of a bathroom has a toilet and a corner in it .",
"a corner has . bathroom a in toilet of A telephone and it a",
"corner A of has bathroom a toilet and a a in telephone . it",
"a telephone in a toilet and a bathroom has A corner of it ."
] |
coco_60072_348513 | A white toilet sitting next to a hair dryer. | datasets/COCO/train2014/COCO_train2014_000000290361.jpg | [
"white toilet",
"hair dryer"
] | [
{
"positive": "white toilet",
"negative": "black toilet",
"change_type": "attribute_change"
},
{
"positive": "white toilet",
"negative": "white sink",
"change_type": "object_change"
},
{
"positive": "hair dryer",
"negative": "toothbrush",
"change_type": "object_change"
... | [{"subject": "white toilet", "relation_type": "is sitting next to", "object": "hair dryer", "negatives": [{"subject": "white toilet", "relation_type": "is away from", "object": "hair dryer", "change_type": "antonym"}, {"subject": "white toilet", "relation_type": "is not sitting next to", "object": "hair dryer", "change... | [
"A toilet hair sitting next to a white dryer .",
"A white toilet to a sitting . hair dryer next",
"A toilet white next to sitting dryer hair a .",
"a hair dryer . A white toilet sitting next to"
] |
coco_60072_349764 | A phone is on the wall in this bathroom. | datasets/COCO/train2014/COCO_train2014_000000290361.jpg | [
"phone",
"wall",
"bathroom"
] | [
{
"positive": "phone",
"negative": "mirror",
"change_type": "object_change"
},
{
"positive": "wall",
"negative": "floor",
"change_type": "object_change"
},
{
"positive": "wall",
"negative": "ceiling",
"change_type": "object_change"
},
{
"positive": "bathroom",
... | [{"subject": "phone", "relation_type": "is on", "object": "wall", "negatives": [{"subject": "phone", "relation_type": "is under", "object": "wall", "change_type": "antonym"}, {"subject": "phone", "relation_type": "is not on", "object": "wall", "change_type": "negation"}, {"subject": "wall", "relation_type": "is on", "o... | [
"A bathroom is on the phone in this wall .",
"the phone on this . wall is in bathroom A",
"is phone A the wall on this in bathroom .",
"A phone is . in this bathroom on the wall"
] |
coco_60072_354723 | A toilet with a flushing system and a wall telephone. | datasets/COCO/train2014/COCO_train2014_000000290361.jpg | [
"toilet",
"flushing system",
"wall telephone"
] | [
{
"positive": "toilet",
"negative": "sink",
"change_type": "object_change"
},
{
"positive": "flushing system",
"negative": "washing machine",
"change_type": "object_change"
},
{
"positive": "wall telephone",
"negative": "wall light",
"change_type": "object_change"
},
... | [{"subject": "toilet", "relation_type": "has", "object": "flushing system", "negatives": [{"subject": "toilet", "relation_type": "does not have", "object": "flushing system", "change_type": "negation"}, {"subject": "toilet", "relation_type": "is detached from", "object": "flushing system", "change_type": "antonym"}]}, ... | [
"A system with a flushing wall and a telephone toilet .",
". toilet a and flushing system with A wall telephone a",
"toilet with A system a flushing and wall a . telephone",
"telephone . A toilet with a flushing system and a wall"
] |
coco_60072_356250 | A toilet in a bathroom that has a phone. | datasets/COCO/train2014/COCO_train2014_000000290361.jpg | [
"toilet",
"bathroom",
"phone"
] | [
{
"positive": "toilet",
"negative": "sink",
"change_type": "object_change"
},
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "phone",
"negative": "toaster",
"change_type": "object_change"
},
{
"positive": "phone",
... | [{"subject": "toilet", "relation_type": "is in", "object": "bathroom", "negatives": [{"subject": "toilet", "relation_type": "is outside", "object": "bathroom", "change_type": "antonym"}, {"subject": "toilet", "relation_type": "is not in", "object": "bathroom", "change_type": "negation"}]}, {"subject": "bathroom", "rela... | [
"A bathroom in a toilet that has a phone .",
"A toilet in has bathroom that a a phone .",
"A in toilet that a bathroom a has phone .",
". a bathroom that has a phone A toilet in"
] |
coco_60073_97278 | A bowl of soup is garnished with bacon. | datasets/COCO/train2014/COCO_train2014_000000442489.jpg | [
"bowl of soup",
"bacon"
] | [
{
"positive": "bowl of soup",
"negative": "bowl of salad",
"change_type": "object_change"
},
{
"positive": "bacon",
"negative": "ham",
"change_type": "object_change"
}
] | [{"subject": "bowl of soup", "relation_type": "is garnished with", "object": "bacon", "negatives": [{"subject": "bowl of soup", "relation_type": "is not garnished with", "object": "bacon", "change_type": "negation"}, {"subject": "bowl of soup", "relation_type": "is stripped of", "object": "bacon", "change_type": "anton... | [
"A soup of bowl is garnished with bacon .",
"A bowl garnished soup with . is bacon of",
"of bowl A garnished soup is . bacon with",
"with bacon . soup is garnished A bowl of"
] |
coco_60073_99900 | A BOWL OF SOUP THAT IS SITTING ON THE TABLE | datasets/COCO/train2014/COCO_train2014_000000442489.jpg | [
"a bowl of soup",
"the table"
] | [
{
"positive": "a bowl of soup",
"negative": "a plate of salad",
"change_type": "object_change"
},
{
"positive": "a bowl of soup",
"negative": "a bowl of cereal",
"change_type": "object_change"
},
{
"positive": "the table",
"negative": "the chair",
"change_type": "object_c... | [{"subject": "a bowl of soup", "relation_type": "is sitting on", "object": "the table", "negatives": [{"subject": "a bowl of soup", "relation_type": "is standing on", "object": "the table", "change_type": "antonym"}, {"subject": "a bowl of soup", "relation_type": "is under", "object": "the table", "change_type": "anton... | [
"A BOWL THE SOUP SITTING OF ON THAT IS TABLE",
"A OF BOWL IS THAT SOUP SITTING ON THE TABLE",
"TABLE SOUP THAT IS A BOWL OF SITTING ON THE"
] |
coco_60073_100401 | A spoon rests in a bowl of soup. | datasets/COCO/train2014/COCO_train2014_000000442489.jpg | [
"spoon",
"bowl of soup"
] | [
{
"positive": "spoon",
"negative": "fork",
"change_type": "object_change"
},
{
"positive": "bowl of soup",
"negative": "plate of pasta",
"change_type": "object_change"
},
{
"positive": "bowl of soup",
"negative": "bowl of salad",
"change_type": "attribute_change"
}
] | [{"subject": "spoon", "relation_type": "is resting in", "object": "bowl of soup", "negatives": [{"subject": "spoon", "relation_type": "is not resting in", "object": "bowl of soup", "change_type": "negation"}, {"subject": "spoon", "relation_type": "is outside", "object": "bowl of soup", "change_type": "antonym"}, {"subj... | [
"A spoon bowl in a rests of soup .",
"A spoon rests a of bowl in soup .",
"A rests spoon bowl in a . of soup",
"of soup . A spoon rests in a bowl"
] |
coco_60073_111966 | A black bowl of soup on a green tray | datasets/COCO/train2014/COCO_train2014_000000442489.jpg | [
"black bowl of soup",
"green tray"
] | [
{
"positive": "black bowl of soup",
"negative": "white bowl of soup",
"change_type": "attribute_change"
},
{
"positive": "black bowl of soup",
"negative": "black bowl of salad",
"change_type": "object_change"
},
{
"positive": "green tray",
"negative": "red tray",
"change_... | [{"subject": "black bowl of soup", "relation_type": "is on", "object": "green tray", "negatives": [{"subject": "black bowl of soup", "relation_type": "is under", "object": "green tray", "change_type": "antonym"}, {"subject": "black bowl of soup", "relation_type": "is beside", "object": "green tray", "change_type": "ant... | [
"A black green of soup on a bowl tray",
"A black bowl on soup a of green tray",
"black A bowl soup on of tray a green",
"A black bowl a green tray of soup on"
] |
coco_60073_117483 | A blue bowl filled with soup and a metal spoon. | datasets/COCO/train2014/COCO_train2014_000000442489.jpg | [
"blue bowl",
"soup",
"metal spoon"
] | [
{
"positive": "blue bowl",
"negative": "red plate",
"change_type": "object_change"
},
{
"positive": "blue bowl",
"negative": "blue cup",
"change_type": "object_change"
},
{
"positive": "soup",
"negative": "salad",
"change_type": "object_change"
},
{
"positive": "m... | [{"subject": "blue bowl", "relation_type": "is filled with", "object": "soup", "negatives": [{"subject": "blue bowl", "relation_type": "is empty of", "object": "soup", "change_type": "antonym"}, {"subject": "blue bowl", "relation_type": "is not filled with", "object": "soup", "change_type": "negation"}]}, {"subject": "... | [
"A metal bowl filled with blue and a spoon soup .",
"and blue bowl with a soup filled A metal spoon .",
"bowl A blue soup with filled metal a and spoon .",
"spoon . A blue bowl and a metal filled with soup"
] |
coco_60074_589252 | A blue bowl filled with soup on top of a green plate. | datasets/COCO/train2014/COCO_train2014_000000462968.jpg | [
"blue bowl",
"soup",
"green plate"
] | [
{
"positive": "blue bowl",
"negative": "red bowl",
"change_type": "attribute_change"
},
{
"positive": "blue bowl",
"negative": "blue cup",
"change_type": "object_change"
},
{
"positive": "soup",
"negative": "salad",
"change_type": "object_change"
},
{
"positive": ... | [{"subject": "blue bowl", "relation_type": "is filled with", "object": "soup", "negatives": [{"subject": "blue bowl", "relation_type": "is empty of", "object": "soup", "change_type": "antonym"}, {"subject": "blue bowl", "relation_type": "is not filled with", "object": "soup", "change_type": "negation"}]}, {"subject": "... | [
"A soup blue filled with green on bowl of a plate top .",
"a blue bowl with filled soup of top A . green plate on",
"blue A bowl soup filled with on of top plate green a .",
"a green plate . A blue bowl on top of filled with soup"
] |
coco_60074_590521 | A plate of soup that looks to have broccoli in it as well as many other items. | datasets/COCO/train2014/COCO_train2014_000000462968.jpg | [
"plate of soup",
"broccoli",
"many other items"
] | [
{
"positive": "plate of soup",
"negative": "plate of salad",
"change_type": "object_change"
},
{
"positive": "broccoli",
"negative": "carrot",
"change_type": "object_change"
}
] | [{"subject": "plate of soup", "relation_type": "contains", "object": "broccoli", "negatives": []}, {"subject": "plate of soup", "relation_type": "contains", "object": "many other items", "negatives": []}] | [
"A soup of plate that looks to have many in it as well as items other broccoli .",
"well plate . soup looks in of to broccoli it that A as as many other items have",
"of plate A that looks soup broccoli to have as in it well many as other items .",
"to have broccoli in it as soup that looks A plate of well as... |
coco_60074_592732 | Creme soup or chowder with broccoli and bacon in a blue bowl with a spoon, on a green plate, on a white cloth. | datasets/COCO/train2014/COCO_train2014_000000462968.jpg | [
"creme soup or chowder",
"broccoli",
"bacon",
"blue bowl",
"spoon",
"green plate",
"white cloth"
] | [
{
"positive": "creme soup or chowder",
"negative": "chili soup with carrots and sausage in a red bowl with a fork, on a yellow plate, on a brown tablecloth",
"change_type": "object_change"
},
{
"positive": "broccoli",
"negative": "carrot",
"change_type": "object_change"
},
{
"pos... | [{"subject": "creme soup or chowder", "relation_type": "is in", "object": "blue bowl", "negatives": [{"subject": "creme soup or chowder", "relation_type": "is not in", "object": "blue bowl", "change_type": "negation"}, {"subject": "creme soup or chowder", "relation_type": "is outside", "object": "blue bowl", "change_ty... | [
"broccoli plate or cloth with Creme and white in a chowder bacon with a spoon , on a blue soup , on a bowl green .",
"Creme soup in chowder with broccoli or bacon a . blue bowl on and spoon a a , green plate a on , white cloth with",
"or Creme soup with broccoli chowder in bacon and bowl a blue a with spoon a ,... |
coco_60074_593008 | a bowl full of some kind of soup | datasets/COCO/train2014/COCO_train2014_000000462968.jpg | [
"bowl",
"soup"
] | [
{
"positive": "bowl",
"negative": "plate",
"change_type": "object_change"
},
{
"positive": "soup",
"negative": "salad",
"change_type": "object_change"
}
] | [{"subject": "bowl", "relation_type": "is filled with", "object": "soup", "negatives": [{"subject": "bowl", "relation_type": "is empty of", "object": "soup", "change_type": "antonym"}, {"subject": "bowl", "relation_type": "is not filled with", "object": "soup", "change_type": "negation"}, {"subject": "soup", "relation_... | [
"a bowl kind of some full of soup",
"some bowl full of a kind of soup",
"of soup a bowl full of some kind"
] |
coco_60074_594718 | a blue bowl is filled with some soup | datasets/COCO/train2014/COCO_train2014_000000462968.jpg | [
"blue bowl",
"soup"
] | [
{
"positive": "blue bowl",
"negative": "red bowl",
"change_type": "attribute_change"
},
{
"positive": "blue bowl",
"negative": "blue plate",
"change_type": "object_change"
},
{
"positive": "soup",
"negative": "a red bowl is filled with some salad",
"change_type": "object_... | [{"subject": "blue bowl", "relation_type": "is filled with", "object": "soup", "negatives": [{"subject": "blue bowl", "relation_type": "is empty of", "object": "soup", "change_type": "antonym"}, {"subject": "blue bowl", "relation_type": "is not filled with", "object": "soup", "change_type": "negation"}, {"subject": "so... | [
"a bowl soup is filled with some blue",
"a blue bowl is filled some with soup",
"blue a bowl filled is with some soup",
"some soup is filled with a blue bowl"
] |
coco_60075_200502 | A bathroom with urinals where women are looking. | datasets/COCO/train2014/COCO_train2014_000000369644.jpg | [
"bathroom",
"urinals",
"women"
] | [
{
"positive": "bathroom",
"negative": "kitchen",
"change_type": "object_change"
},
{
"positive": "urinals",
"negative": "toilets",
"change_type": "object_change"
},
{
"positive": "women",
"negative": "children",
"change_type": "attribute_change"
}
] | [{"subject": "bathroom", "relation_type": "has", "object": "urinals", "negatives": [{"subject": "bathroom", "relation_type": "is without", "object": "urinals", "change_type": "antonym"}, {"subject": "bathroom", "relation_type": "does not have", "object": "urinals", "change_type": "negation"}]}, {"subject": "women", "re... | [
"A bathroom with women where urinals are looking .",
"A bathroom . urinals looking women are with where",
"bathroom A with urinals women where . looking are",
"A bathroom with urinals where women are looking ."
] |
coco_60075_203571 | A row of three wall mounted urinals with a woman peeking behind a tree. | datasets/COCO/train2014/COCO_train2014_000000369644.jpg | [
"row of three wall mounted urinals",
"woman",
"tree"
] | [
{
"positive": "woman",
"negative": "girl",
"change_type": "attribute_change"
},
{
"positive": "tree",
"negative": "bush",
"change_type": "object_change"
}
] | [{"subject": "woman", "relation_type": "is peeking behind", "object": "tree", "negatives": [{"subject": "woman", "relation_type": "is not peeking behind", "object": "tree", "change_type": "negation"}, {"subject": "woman", "relation_type": "is in front of", "object": "tree", "change_type": "antonym"}, {"subject": "tree"... | [
"A woman of three tree mounted urinals with a row peeking behind a wall .",
". row A of wall peeking urinals with behind woman three a a tree mounted",
"row A of mounted three wall a urinals with woman behind peeking a . tree",
"a tree . three wall mounted woman peeking behind urinals with a A row of"
] |
coco_60075_206745 | I line of urinals on a wall decorated with photos of a woman peeking out from behind a tree. | datasets/COCO/train2014/COCO_train2014_000000369644.jpg | [
"line of urinals",
"wall",
"photos of a woman",
"tree"
] | [
{
"positive": "line of urinals",
"negative": "row of sinks",
"change_type": "object_change"
},
{
"positive": "wall",
"negative": "floor",
"change_type": "object_change"
},
{
"positive": "wall",
"negative": "ceiling",
"change_type": "object_change"
},
{
"positive":... | [{"subject": "line of urinals", "relation_type": "is on", "object": "wall", "negatives": [{"subject": "line of urinals", "relation_type": "is under", "object": "wall", "change_type": "antonym"}, {"subject": "line of urinals", "relation_type": "is not on", "object": "wall", "change_type": "negation"}, {"subject": "wall"... | [
"I line of photos on a urinals decorated with wall of a tree peeking out from behind a woman .",
"a peeking . urinals out of wall behind a photos I line woman with from decorated on a tree of",
"of line I on urinals a wall with decorated photos of a peeking out woman from behind a . tree",
"urinals on a tree ... |
coco_60075_206982 | a woman in a black and white dress and three white urinals | datasets/COCO/train2014/COCO_train2014_000000369644.jpg | [
"woman",
"black and white dress",
"three white urinals"
] | [
{
"positive": "black and white dress",
"negative": "black and white shirt",
"change_type": "object_change"
},
{
"positive": "black and white dress",
"negative": "colorful and white dress",
"change_type": "attribute_change"
},
{
"positive": "three white urinals",
"negative": "... | [{"subject": "woman", "relation_type": "is wearing", "object": "black and white dress", "negatives": [{"subject": "woman", "relation_type": "is not wearing", "object": "black and white dress", "change_type": "negation"}, {"subject": "woman", "relation_type": "is detached from", "object": "black and white dress", "chang... | [
"a white in a black and dress urinals and three woman white",
"a woman and three black a white dress in and white urinals",
"in woman a and black a white and dress three white urinals",
"a woman in three white urinals a black and white dress and"
] |
coco_60075_211734 | Several urinals with painted women on the walls. | datasets/COCO/train2014/COCO_train2014_000000369644.jpg | [
"several urinals",
"painted women"
] | [
{
"positive": "several urinals",
"negative": "several sinks",
"change_type": "object_change"
},
{
"positive": "painted women",
"negative": "painted men",
"change_type": "attribute_change"
},
{
"positive": "painted women",
"negative": "painted landscapes",
"change_type": "... | [{"subject": "urinals", "relation_type": "have", "object": "painted women", "negatives": []}] | [
"urinals walls with painted Several on the women .",
"Several urinals . on women the painted walls with",
"Several with urinals painted on women the . walls",
"painted women on the walls . Several urinals with"
] |
coco_60076_4219 | two motorcycles and a car parked on the side of an empty road. | datasets/COCO/train2014/COCO_train2014_000000480014.jpg | [
"two motorcycles",
"car",
"empty road"
] | [
{
"positive": "two motorcycles",
"negative": "one motorcycle",
"change_type": "attribute_change"
},
{
"positive": "car",
"negative": "bus",
"change_type": "object_change"
},
{
"positive": "empty road",
"negative": "crowded road",
"change_type": "attribute_change"
},
{... | [{"subject": "two motorcycles", "relation_type": "are parked on", "object": "empty road", "negatives": [{"subject": "two motorcycles", "relation_type": "are not parked on", "object": "empty road", "change_type": "negation"}, {"subject": "two motorcycles", "relation_type": "are away from", "object": "empty road", "chang... | [
"two motorcycles and a side parked on the car of an empty road .",
"an motorcycles the parked car on of a side two and empty road .",
"and two motorcycles a car parked side on the of empty an road .",
"on the side a car parked two motorcycles and road . of an empty"
] |
coco_60076_6370 | Two motorcycles and a car parked on the curb | datasets/COCO/train2014/COCO_train2014_000000480014.jpg | [
"two motorcycles",
"car",
"curb"
] | [
{
"positive": "two motorcycles",
"negative": "one motorcycle",
"change_type": "attribute_change"
},
{
"positive": "car",
"negative": "bus",
"change_type": "object_change"
},
{
"positive": "curb",
"negative": "sidewalk",
"change_type": "object_change"
},
{
"positiv... | [{"subject": "two motorcycles", "relation_type": "are parked on", "object": "curb", "negatives": [{"subject": "two motorcycles", "relation_type": "are not parked on", "object": "curb", "change_type": "negation"}, {"subject": "two motorcycles", "relation_type": "are away from", "object": "curb", "change_type": "antonym"... | [
"Two car and a motorcycles parked on the curb",
"a motorcycles the and car parked Two on curb",
"and motorcycles Two car a parked the curb on",
"Two motorcycles and on the curb a car parked"
] |
coco_60076_9574 | Motorcycles parked on the side of a street. | datasets/COCO/train2014/COCO_train2014_000000480014.jpg | [
"Motorcycles",
"side of a street"
] | [
{
"positive": "Motorcycles",
"negative": "Bicycles",
"change_type": "object_change"
},
{
"positive": "side of a street",
"negative": "corner of a street",
"change_type": "object_change"
},
{
"positive": "side of a street",
"negative": "middle of a street",
"change_type": ... | [{"subject": "Motorcycles", "relation_type": "are parked on", "object": "side of a street", "negatives": [{"subject": "Motorcycles", "relation_type": "are not parked on", "object": "side of a street", "change_type": "negation"}, {"subject": "Motorcycles", "relation_type": "are parked under", "object": "side of a street... | [
"side parked on the street of a Motorcycles .",
"Motorcycles a on . side the parked street of",
"parked on Motorcycles side of the a street .",
"a street . Motorcycles parked on the side of"
] |
coco_60076_11776 | A city street at night, with parked motorcycles and a sunset in the distance. | datasets/COCO/train2014/COCO_train2014_000000480014.jpg | [
"city street",
"parked motorcycles",
"sunset"
] | [
{
"positive": "city street",
"negative": "forest path",
"change_type": "object_change"
},
{
"positive": "parked motorcycles",
"negative": "parked cars",
"change_type": "object_change"
},
{
"positive": "parked motorcycles",
"negative": "parked bicycles",
"change_type": "ob... | [{"subject": "city street", "relation_type": "has", "object": "parked motorcycles", "negatives": [{"subject": "city street", "relation_type": "does not have", "object": "parked motorcycles", "change_type": "negation"}, {"subject": "city street", "relation_type": "is without", "object": "parked motorcycles", "change_typ... | [
"A street night at city , with parked sunset and a distance in the motorcycles .",
"with city street the night , and . motorcycles in a sunset at A distance parked",
"A street city , at night motorcycles with parked and sunset a the in distance .",
"and a sunset at night , with parked motorcycles in the dista... |
coco_60076_819686 | Two motorcycles parked on a street at night. | datasets/COCO/train2014/COCO_train2014_000000480014.jpg | [
"two motorcycles",
"street",
"night"
] | [
{
"positive": "two motorcycles",
"negative": "one motorcycle",
"change_type": "attribute_change"
},
{
"positive": "two motorcycles",
"negative": "two cars",
"change_type": "object_change"
},
{
"positive": "street",
"negative": "forest",
"change_type": "object_change"
},... | [{"subject": "two motorcycles", "relation_type": "are parked on", "object": "street", "negatives": [{"subject": "two motorcycles", "relation_type": "are not parked on", "object": "street", "change_type": "negation"}, {"subject": "two motorcycles", "relation_type": "are moving away from", "object": "street", "change_typ... | [
"Two night parked on a motorcycles at street .",
"on motorcycles parked at . street Two night a",
"Two motorcycles parked a on street night at .",
"Two motorcycles parked at night . on a street"
] |
coco_60077_575452 | A cat laying in a white bowl on top of a cupboard. | datasets/COCO/train2014/COCO_train2014_000000534722.jpg | [
"cat",
"white bowl",
"cupboard"
] | [
{
"positive": "cat",
"negative": "dog",
"change_type": "object_change"
},
{
"positive": "white bowl",
"negative": "black bowl",
"change_type": "attribute_change"
},
{
"positive": "white bowl",
"negative": "white plate",
"change_type": "object_change"
},
{
"positiv... | [{"subject": "cat", "relation_type": "is laying in", "object": "white bowl", "negatives": [{"subject": "cat", "relation_type": "is not laying in", "object": "white bowl", "change_type": "negation"}, {"subject": "cat", "relation_type": "is standing in", "object": "white bowl", "change_type": "antonym"}, {"subject": "cat... | [
"A white laying in a cupboard cat on bowl of a top .",
"laying cat A . a white bowl a top on in cupboard of",
"A laying cat in a white top on bowl a cupboard of .",
"in a white . A cat laying of a cupboard bowl on top"
] |
coco_60077_576733 | a cat laying in a bowl on top of a cabinet | datasets/COCO/train2014/COCO_train2014_000000534722.jpg | [
"cat",
"bowl",
"cabinet"
] | [
{
"positive": "cat",
"negative": "dog",
"change_type": "object_change"
},
{
"positive": "bowl",
"negative": "plate",
"change_type": "object_change"
},
{
"positive": "cabinet",
"negative": "shelf",
"change_type": "object_change"
}
] | [{"subject": "cat", "relation_type": "is laying in", "object": "bowl", "negatives": [{"subject": "cat", "relation_type": "is not laying in", "object": "bowl", "change_type": "negation"}, {"subject": "cat", "relation_type": "is standing in", "object": "bowl", "change_type": "antonym"}, {"subject": "cat", "relation_type"... | [
"a top laying in a cabinet on bowl of a cat",
"in cat a laying of bowl a top a on cabinet",
"a cat laying in bowl a of on top a cabinet",
"a cat laying in a bowl a cabinet on top of"
] |
coco_60077_579613 | A cat looking down from inside of a bowl on top of a cabinet. | datasets/COCO/train2014/COCO_train2014_000000534722.jpg | [
"cat",
"bowl",
"cabinet"
] | [
{
"positive": "cat",
"negative": "dog",
"change_type": "object_change"
},
{
"positive": "bowl",
"negative": "plate",
"change_type": "object_change"
},
{
"positive": "cabinet",
"negative": "shelf",
"change_type": "object_change"
}
] | [{"subject": "cat", "relation_type": "is looking down from inside of", "object": "bowl", "negatives": [{"subject": "cat", "relation_type": "is looking up from inside of", "object": "bowl", "change_type": "antonym"}, {"subject": "cat", "relation_type": "is not looking down from inside of", "object": "bowl", "change_type... | [
"A top looking down from inside of a cat on cabinet of a bowl .",
"of cat inside looking a . of A bowl on top from a cabinet down",
"cat A looking inside down from a bowl of top of on a . cabinet",
"A cat looking down from inside on top of of a bowl a cabinet ."
] |
coco_60077_583321 | A cat sits in a bowl placed on top of a cabinet. | datasets/COCO/train2014/COCO_train2014_000000534722.jpg | [
"cat",
"bowl",
"cabinet"
] | [
{
"positive": "cat",
"negative": "dog",
"change_type": "object_change"
},
{
"positive": "bowl",
"negative": "plate",
"change_type": "object_change"
},
{
"positive": "cabinet",
"negative": "shelf",
"change_type": "object_change"
}
] | [{"subject": "cat", "relation_type": "is sitting in", "object": "bowl", "negatives": [{"subject": "cat", "relation_type": "is not sitting in", "object": "bowl", "change_type": "negation"}, {"subject": "cat", "relation_type": "is standing in", "object": "bowl", "change_type": "antonym"}, {"subject": "cat", "relation_typ... | [
"A top sits in a cat placed on cabinet of a bowl .",
". cat of on sits bowl a placed top A a cabinet in",
"cat sits A in a bowl top on placed of a cabinet .",
"in a bowl . placed on top of a cabinet A cat sits"
] |
coco_60077_587845 | a cat laying in a bowl on top of a cabinet | datasets/COCO/train2014/COCO_train2014_000000534722.jpg | [
"cat",
"bowl",
"cabinet"
] | [
{
"positive": "cat",
"negative": "dog",
"change_type": "object_change"
},
{
"positive": "bowl",
"negative": "plate",
"change_type": "object_change"
},
{
"positive": "cabinet",
"negative": "shelf",
"change_type": "object_change"
}
] | [{"subject": "cat", "relation_type": "is laying in", "object": "bowl", "negatives": [{"subject": "cat", "relation_type": "is not laying in", "object": "bowl", "change_type": "negation"}, {"subject": "cat", "relation_type": "is standing in", "object": "bowl", "change_type": "antonym"}, {"subject": "cat", "relation_type"... | [
"a cabinet laying in a bowl on top of a cat",
"in cat a on a bowl laying top of a cabinet",
"cat laying a bowl in a on of top cabinet a",
"a cabinet on top of in a bowl a cat laying"
] |
coco_60078_695605 | Several different computers are next to each other on a desk. | datasets/COCO/train2014/COCO_train2014_000000474472.jpg | [
"several different computers",
"desk"
] | [
{
"positive": "several different computers",
"negative": "several different phones",
"change_type": "object_change"
},
{
"positive": "several different computers",
"negative": "two different computers",
"change_type": "attribute_change"
},
{
"positive": "desk",
"negative": "t... | [{"subject": "several different computers", "relation_type": "are next to each other on", "object": "desk", "negatives": [{"subject": "several different computers", "relation_type": "are away from each other on", "object": "desk", "change_type": "antonym"}, {"subject": "several different computers", "relation_type": "a... | [
"different other next are computers to each Several on a desk .",
"Several different computers a next to each other are . desk on",
"different computers Several to are next each other on a . desk",
"Several different computers each other on a desk . are next to"
] |
coco_60078_699547 | A computer workstation set up with several monitors and laptops. | datasets/COCO/train2014/COCO_train2014_000000474472.jpg | [
"computer workstation",
"several monitors",
"laptops"
] | [
{
"positive": "computer workstation",
"negative": "office desk",
"change_type": "object_change"
},
{
"positive": "computer workstation",
"negative": "home entertainment center",
"change_type": "object_change"
},
{
"positive": "several monitors",
"negative": "one monitor",
... | [{"subject": "computer workstation", "relation_type": "is set up with", "object": "several monitors", "negatives": [{"subject": "computer workstation", "relation_type": "is not set up with", "object": "several monitors", "change_type": "negation"}, {"subject": "computer workstation", "relation_type": "is disconnected f... | [
"A computer workstation set up with monitors laptops and several .",
"and computer workstation set up with several monitors A laptops .",
"computer workstation A up with set several and monitors laptops .",
"several monitors and A computer workstation laptops . set up with"
] |
coco_60078_705244 | a bunch of computers on a desk together. | datasets/COCO/train2014/COCO_train2014_000000474472.jpg | [
"a bunch of computers",
"desk"
] | [
{
"positive": "a bunch of computers",
"negative": "a bunch of phones",
"change_type": "object_change"
},
{
"positive": "a bunch of computers",
"negative": "a bunch of laptops",
"change_type": "object_change"
},
{
"positive": "a bunch of computers",
"negative": "a bunch of tab... | [{"subject": "a bunch of computers", "relation_type": "are on", "object": "desk", "negatives": [{"subject": "a bunch of computers", "relation_type": "are under", "object": "desk", "change_type": "antonym"}, {"subject": "a bunch of computers", "relation_type": "are not on", "object": "desk", "change_type": "negation"}, ... | [
"a computers of desk on a bunch together .",
"a bunch a computers . together desk on of",
"bunch a of computers a on . together desk",
"desk together . computers on a a bunch of"
] |
coco_60078_705340 | There are four monitors and two keyboards on the desk. | datasets/COCO/train2014/COCO_train2014_000000474472.jpg | [
"four monitors",
"two keyboards",
"desk"
] | [
{
"positive": "four monitors",
"negative": "two monitors",
"change_type": "attribute_change"
},
{
"positive": "four monitors",
"negative": "four laptops",
"change_type": "object_change"
},
{
"positive": "two keyboards",
"negative": "one keyboard",
"change_type": "attribut... | [{"subject": "four monitors", "relation_type": "are on", "object": "desk", "negatives": [{"subject": "four monitors", "relation_type": "are under", "object": "desk", "change_type": "antonym"}, {"subject": "four monitors", "relation_type": "are not on", "object": "desk", "change_type": "negation"}, {"subject": "desk", "... | [
"There are four desk and two keyboards on the monitors .",
"two on There monitors are four keyboards and the desk .",
"are There four two and monitors the on keyboards . desk",
"There are four desk . monitors and two keyboards on the"
] |
coco_60078_705814 | Several computers lined up on one desk next to each other. | datasets/COCO/train2014/COCO_train2014_000000474472.jpg | [
"Several computers",
"desk"
] | [
{
"positive": "Several computers",
"negative": "Several laptops",
"change_type": "object_change"
},
{
"positive": "Several computers",
"negative": "Two computers",
"change_type": "attribute_change"
},
{
"positive": "desk",
"negative": "table",
"change_type": "object_chang... | [{"subject": "Several computers", "relation_type": "are lined up on", "object": "desk", "negatives": [{"subject": "Several computers", "relation_type": "are not lined up on", "object": "desk", "change_type": "negation"}, {"subject": "desk", "relation_type": "are lined up on", "object": "Several computers", "change_type... | [
"desk computers lined up on one Several next to each other .",
"Several computers . one to up desk on each next other lined",
"computers lined Several on up one to desk next each other .",
"each other . up on one desk next to Several computers lined"
] |
coco_60079_14288 | a still life painting of a bowl of fruit and a tea set on a tablecloth | datasets/COCO/train2014/COCO_train2014_000000039315.jpg | [
"still life painting",
"bowl of fruit",
"tea set",
"tablecloth"
] | [
{
"positive": "bowl of fruit",
"negative": "bowl of vegetables",
"change_type": "object_change"
},
{
"positive": "tea set",
"negative": "coffee set",
"change_type": "object_change"
}
] | [{"subject": "still life painting", "relation_type": "contains", "object": "bowl of fruit", "negatives": [{"subject": "still life painting", "relation_type": "does not contain", "object": "bowl of fruit", "change_type": "negation"}, {"subject": "still life painting", "relation_type": "is separate from", "object": "bowl... | [
"a bowl still painting of a life of fruit and a tea set on a tablecloth",
"of still life painting and a bowl a fruit set a tea on of a tablecloth",
"a still life of a painting fruit bowl of tea a and on set a tablecloth",
"bowl of fruit and a tea tablecloth a still life painting of a set on a"
] |
coco_60079_16559 | A painting of a cloth, a teapot, sugar, and three oranges. | datasets/COCO/train2014/COCO_train2014_000000039315.jpg | [
"painting",
"cloth",
"teapot",
"sugar",
"three oranges"
] | [
{
"positive": "painting",
"negative": "drawing",
"change_type": "object_change"
},
{
"positive": "painting",
"negative": "sculpture",
"change_type": "object_change"
},
{
"positive": "cloth",
"negative": "fabric",
"change_type": "attribute_change"
},
{
"positive": ... | [{"subject": "painting", "relation_type": "is of", "object": "cloth", "negatives": [{"subject": "painting", "relation_type": "is not of", "object": "cloth", "change_type": "negation"}]}, {"subject": "painting", "relation_type": "is of", "object": "teapot", "negatives": [{"subject": "painting", "relation_type": "is not ... | [
"A painting of a sugar , a teapot , cloth , and three oranges .",
"a painting a and cloth , of teapot three sugar A , , oranges .",
"A of painting a cloth , , teapot a and sugar , three . oranges",
"three oranges . a cloth , a teapot , A painting of sugar , and"
] |
coco_60079_19445 | A painting with oranges and dishware on a tablecloth. | datasets/COCO/train2014/COCO_train2014_000000039315.jpg | [
"painting",
"oranges",
"dishware",
"tablecloth"
] | [
{
"positive": "painting",
"negative": "sculpture",
"change_type": "object_change"
},
{
"positive": "oranges",
"negative": "apples",
"change_type": "object_change"
},
{
"positive": "dishware",
"negative": "cutlery",
"change_type": "object_change"
},
{
"positive": "... | [{"subject": "painting", "relation_type": "has", "object": "oranges", "negatives": [{"subject": "painting", "relation_type": "does not have", "object": "oranges", "change_type": "negation"}, {"subject": "painting", "relation_type": "is without", "object": "oranges", "change_type": "antonym"}]}, {"subject": "painting", ... | [
"A tablecloth with oranges and dishware on a painting .",
"a painting dishware oranges with . on A tablecloth and",
"painting A with and oranges dishware tablecloth on a .",
"A painting with on a tablecloth . oranges and dishware"
] |
coco_60079_20810 | a painting of a blanket with three oranges | datasets/COCO/train2014/COCO_train2014_000000039315.jpg | [
"painting",
"blanket",
"three oranges"
] | [
{
"positive": "three oranges",
"negative": "one orange",
"change_type": "attribute_change"
},
{
"positive": "three oranges",
"negative": "three apples",
"change_type": "object_change"
}
] | [{"subject": "painting", "relation_type": "is of", "object": "blanket", "negatives": [{"subject": "painting", "relation_type": "is not of", "object": "blanket", "change_type": "negation"}]}, {"subject": "painting", "relation_type": "includes", "object": "three oranges", "negatives": [{"subject": "painting", "relation_t... | [
"three painting with a blanket a of oranges",
"a of painting blanket a with three oranges",
"three oranges a blanket with a painting of"
] |
coco_60079_31040 | A portrait of a teapot and oranges on a saucer. | datasets/COCO/train2014/COCO_train2014_000000039315.jpg | [
"portrait",
"teapot",
"oranges",
"saucer"
] | [
{
"positive": "teapot",
"negative": "coffee mug",
"change_type": "object_change"
},
{
"positive": "oranges",
"negative": "apples",
"change_type": "object_change"
},
{
"positive": "saucer",
"negative": "plate",
"change_type": "object_change"
}
] | [{"subject": "portrait", "relation_type": "includes", "object": "teapot", "negatives": [{"subject": "portrait", "relation_type": "excludes", "object": "teapot", "change_type": "antonym"}, {"subject": "portrait", "relation_type": "does not include", "object": "teapot", "change_type": "negation"}]}, {"subject": "portrait... | [
"A portrait of a saucer and oranges on a teapot .",
"of portrait a on teapot A and a oranges saucer .",
"of portrait A teapot a and on oranges a saucer .",
"a teapot and A portrait of saucer . oranges on a"
] |
coco_60080_128921 | A motorcycle with spikes on the wheels on display | datasets/COCO/train2014/COCO_train2014_000000250926.jpg | [
"motorcycle",
"spikes on the wheels",
"display"
] | [
{
"positive": "motorcycle",
"negative": "car",
"change_type": "object_change"
},
{
"positive": "spikes on the wheels",
"negative": "spikes on the helmet",
"change_type": "attribute_change"
},
{
"positive": "spikes on the wheels",
"negative": "spikes on the handlebars",
"c... | [{"subject": "motorcycle", "relation_type": "has", "object": "spikes on the wheels", "negatives": [{"subject": "motorcycle", "relation_type": "is without", "object": "spikes on the wheels", "change_type": "antonym"}, {"subject": "motorcycle", "relation_type": "is not having", "object": "spikes on the wheels", "change_t... | [
"A motorcycle with display on the wheels on spikes",
"the motorcycle A spikes with on wheels on display",
"with motorcycle A spikes the on on display wheels",
"spikes on the A motorcycle with wheels on display"
] |
coco_60080_142844 | Motorcycle with flames painted on it on display | datasets/COCO/train2014/COCO_train2014_000000250926.jpg | [
"motorcycle",
"flames painted on it",
"display"
] | [
{
"positive": "motorcycle",
"negative": "car",
"change_type": "object_change"
},
{
"positive": "flames painted on it",
"negative": "stars painted on it",
"change_type": "attribute_change"
},
{
"positive": "flames painted on it",
"negative": "flowers painted on it",
"chang... | [{"subject": "flames painted on it", "relation_type": "is on", "object": "motorcycle", "negatives": []}, {"subject": "motorcycle", "relation_type": "is on", "object": "display", "negatives": []}] | [
"flames with Motorcycle painted on it on display",
"Motorcycle painted flames on on with it display",
"flames Motorcycle with on painted it on display",
"painted on it Motorcycle with flames on display"
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.