Upload croissant.json

croissant.json

@@ -0,0 +1,408 @@
+{
+    "@context": {
+        "@language": "en",
+        "@vocab": "https://schema.org/",
+        "arrayShape": "cr:arrayShape",
+        "citeAs": "cr:citeAs",
+        "column": "cr:column",
+        "conformsTo": "dct:conformsTo",
+        "containedIn": "cr:containedIn",
+        "cr": "http://mlcommons.org/croissant/",
+        "data": {
+            "@id": "cr:data",
+            "@type": "@json"
+        },
+        "dataBiases": "cr:dataBiases",
+        "dataCollection": "cr:dataCollection",
+        "dataType": {
+            "@id": "cr:dataType",
+            "@type": "@vocab"
+        },
+        "dct": "http://purl.org/dc/terms/",
+        "extract": "cr:extract",
+        "field": "cr:field",
+        "fileProperty": "cr:fileProperty",
+        "fileObject": "cr:fileObject",
+        "fileSet": "cr:fileSet",
+        "format": "cr:format",
+        "includes": "cr:includes",
+        "isArray": "cr:isArray",
+        "isLiveDataset": "cr:isLiveDataset",
+        "jsonPath": "cr:jsonPath",
+        "key": "cr:key",
+        "md5": "cr:md5",
+        "parentField": "cr:parentField",
+        "path": "cr:path",
+        "personalSensitiveInformation": "cr:personalSensitiveInformation",
+        "recordSet": "cr:recordSet",
+        "references": "cr:references",
+        "regex": "cr:regex",
+        "repeated": "cr:repeated",
+        "replace": "cr:replace",
+        "sc": "https://schema.org/",
+        "separator": "cr:separator",
+        "source": "cr:source",
+        "subField": "cr:subField",
+        "transform": "cr:transform",
+        "rai": "http://mlcommons.org/croissant/RAI/",
+        "prov": "http://www.w3.org/ns/prov#"
+    },
+    "@type": "sc:Dataset",
+    "distribution": [
+        {
+            "@type": "cr:FileObject",
+            "@id": "repo",
+            "name": "repo",
+            "description": "The Hugging Face git repository.",
+            "contentUrl": "https://huggingface.co/datasets/ynklab/vocalgrad/tree/refs%2Fconvert%2Fparquet",
+            "encodingFormat": "git+https",
+            "sha256": "https://github.com/mlcommons/croissant/issues/80"
+        },
+        {
+            "@type": "cr:FileSet",
+            "@id": "parquet-files-for-config-default",
+            "containedIn": {
+                "@id": "repo"
+            },
+            "encodingFormat": "application/x-parquet",
+            "includes": "default/*/*.parquet"
+        }
+    ],
+    "recordSet": [
+        {
+            "@type": "cr:RecordSet",
+            "dataType": "cr:Split",
+            "key": {
+                "@id": "default_splits/split_name"
+            },
+            "@id": "default_splits",
+            "name": "default_splits",
+            "description": "Splits for the default config.",
+            "field": [
+                {
+                    "@type": "cr:Field",
+                    "@id": "default_splits/split_name",
+                    "dataType": "sc:Text"
+                }
+            ],
+            "data": [
+                {
+                    "default_splits/split_name": "test"
+                }
+            ]
+        },
+        {
+            "@type": "cr:RecordSet",
+            "@id": "default",
+            "description": "ynklab/vocalgrad - 'default' subset",
+            "field": [
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/split",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "fileProperty": "fullpath"
+                        },
+                        "transform": {
+                            "regex": "default/(?:partial-)?(test)/.+parquet$"
+                        }
+                    },
+                    "references": {
+                        "field": {
+                            "@id": "default_splits/split_name"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/audio",
+                    "dataType": "sc:AudioObject",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "audio"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/category",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "category"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/benchmark_clip_id",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "benchmark_clip_id"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/source_text",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "source_text"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/attribute",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "attribute"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/prompt",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "prompt"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/label",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "label"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/direction",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "direction"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/curve",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "curve"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/tier",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "tier"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/source_clip_id",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "source_clip_id"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/speaker_id",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "speaker_id"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/sentence_id",
+                    "dataType": "cr:Int64",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "sentence_id"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/duration_sec",
+                    "dataType": "cr:Float64",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "duration_sec"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/onset_sec",
+                    "dataType": "cr:Float64",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "onset_sec"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/offset_sec",
+                    "dataType": "cr:Float64",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "offset_sec"
+                        }
+                    }
+                },
+                {
+                    "@type": "cr:Field",
+                    "@id": "default/source_split",
+                    "dataType": "sc:Text",
+                    "source": {
+                        "fileSet": {
+                            "@id": "parquet-files-for-config-default"
+                        },
+                        "extract": {
+                            "column": "split"
+                        }
+                    }
+                }
+            ]
+        }
+    ],
+    "conformsTo": "http://mlcommons.org/croissant/1.1",
+    "name": "vocalgrad",
+    "description": "\n\t\n\t\t\n\t\n\t\n\t\tVocalGrad\n\t\n\nVocalGrad is an audio benchmark for evaluating whether a model can detect the\ndirection of gradual perceptual change in speech. This public release contains\nthe test split only.\nEach example contains one audio clip and one target attribute. The task is to\nanswer whether that attribute increases or decreases over time.\nCode: ynklab/vocalgrad.\n\n\t\n\t\t\n\t\n\t\n\t\tTask\n\t\n\nGiven an audio clip and an attribute name, predict one of two labels:\n\nincrease\ndecrease\n\nThe ground-truth… See the full description on the dataset page: https://huggingface.co/datasets/ynklab/vocalgrad.",
+    "alternateName": [
+        "ynklab/vocalgrad"
+    ],
+    "creator": {
+        "@type": "Organization",
+        "name": "Yanaka Lab.",
+        "url": "https://huggingface.co/ynklab"
+    },
+    "keywords": [
+        "cc-by-4.0",
+        "🇺🇸 Region: US"
+    ],
+    "license": "https://choosealicense.com/licenses/cc-by-4.0/",
+    "url": "https://huggingface.co/datasets/ynklab/vocalgrad",
+    "rai:dataLimitations": "This dataset is designed to evaluate the acoustic perception ability of audio language models under controlled acoustic transformations. However, the set of transformation categories does not necessarily cover all primitive acoustic features relevant to real-world audio understanding.\n\nIn addition, all data points are generated from a fixed set of source clips using a unified augmentation pipeline. As a result, the dataset may contain systematic artifacts introduced by the generation process, and models may overfit to these patterns rather than general acoustic properties.\n\nDue to its synthetic and controlled nature, this dataset is not intended for evaluating performance on natural audio understanding tasks, such as real-world speech recognition, emotion recognition, or environmental sound analysis.",
+    "rai:dataBiases": "The source clips are sampled from the VCTK dataset with an effort to balance speaker attributes such as accent and gender. However, as VCTK is primarily composed of English speakers with specific demographic distributions, residual biases related to language, accent diversity, and recording conditions may remain.\n\nThe primary objective of VocalGrad is to evaluate acoustic perception of controlled temporal changes applied to speech signals, rather than linguistic or demographic attributes. Since each sample is generated through systematic transformations of the same source clips, the influence of demographic factors on task difficulty is expected to be limited.\n\nNevertheless, models may still exhibit different behaviors depending on the underlying speech characteristics (e.g., speaker timbre or recording conditions), and such effects are not explicitly controlled in this benchmark.",
+    "rai:personalSensitiveInformation": "The dataset is derived from the VCTK corpus, which contains speech recordings from human speakers and includes associated attributes such as gender, age, and accent (language-related variation). Therefore, these types of personal information are implicitly present in the source data.\n\nVocalGrad does not directly include these demographic metadata attributes, but the released dataset includes speaker IDs and source clip IDs that can be linked to the original VCTK recordings and associated metadata, as well as source transcripts, transformed audio clips, and derived labels.\n\nVocalGrad does not intentionally collect or annotate sensitive attributes such as health or political information. However, the use of human speech recordings and linkable source identifiers carries a risk of speaker identification inherited from the original dataset.",
+    "rai:dataUseCases": "This dataset is designed to measure the ability of audio language models to perceive controlled temporal changes in basic acoustic features, such as pitch, loudness, and noise. It is validated for benchmarking low-level acoustic perception under controlled conditions, as demonstrated by the clear performance gap between human annotators and state-of-the-art models.\n\nThe dataset is not intended for evaluating performance on high-level real-world tasks, such as emotion recognition, speech understanding, or environmental sound classification. While improvements on this benchmark may provide insights into bottlenecks in higher-level audio understanding, such relationships have not been fully validated and may not directly transfer to real-world settings.",
+    "rai:dataSocialImpact": "VocalGrad is a benchmark for evaluating acoustic perception in audio language models, which may contribute to the development of more reliable and robust audio-based AI systems.\n\nPotential risks arise from the use of real human speech recordings inherited from the VCTK dataset. These include a minimal risk of speaker identification and potential biases related to speaker characteristics such as accent and recording conditions. \n\nTo mitigate these risks, VocalGrad consists only of controlled acoustic transformations and does not introduce additional personal or sensitive information beyond the original dataset.",
+    "rai:hasSyntheticData": true,
+    "prov:wasDerivedFrom": [
+        {
+            "@id": "https://datashare.ed.ac.uk/handle/10283/3443",
+            "prov:label": "VCTK",
+            "sc:license": "CC-BY 4.0"
+        }
+    ],
+    "prov:wasGeneratedBy": [
+        {
+            "@type": "prov:Activity",
+            "prov:type": {
+                "@id": "https://www.wikidata.org/wiki/Q4929239"
+            },
+            "prov:label": "Source clip selection from VCTK",
+            "sc:description": "We select 50 speakers from the VCTK dataset to ensure diversity in accent and gender. For each speaker, we randomly sample two clips that satisfy the filtering criteria, resulting in a total of 100 source clips used for subsequent data processing."
+        },
+        {
+            "@type": "prov:Activity",
+            "prov:type": {
+                "@id": "https://www.wikidata.org/wiki/Q5227332"
+            },
+            "prov:label": "Temporal Change Augmentation",
+            "sc:description": "We detect speech onset and offset timestamps using an RMS-based method and apply nine types of transformations that introduce temporally varying acoustic changes within each clip. Detailed descriptions of each augmentation are provided in the repository README."
+        },
+        {
+            "@type": "prov:Activity",
+            "prov:type": {
+                "@id": "https://www.wikidata.org/wiki/Q1172378"
+            },
+            "prov:label": "Filtering Criteria",
+            "sc:description": "During source clip selection, we filter out clips shorter than three seconds and manually verify the detected onset and offset using waveform visualizations."
+        },
+        {
+            "@type": "prov:Activity",
+            "prov:type": {
+                "@id": "https://www.wikidata.org/wiki/Q3306762"
+            },
+            "prov:label": "Human Perceptibility Verification",
+            "sc:description": "All transformation categories are verified for perceptibility by human annotators to ensure that the applied acoustic changes are reliably detectable."
+        }
+    ]
+}