ImageNet/README.md

PAPE — ImageNet

Real ImageNet photographs edited by four image-editing models. Each manifest row is one language-specific pair: the source image, a semantically normal replacement, a semantically abnormal replacement of the same object, and a binary mask for each edit. Every pair in this folder has normal/abnormal mask IoU of at least 0.5.

2,500 distinct ImageNet image ids are included. A source image can appear more than once when several objects, anomaly categories, or languages were edited.

Counts

Editor	Directory	Pairs	Chinese	English
Qwen-Image-Edit-2511	qwen_image_edit_2511	4,725	2,362	2,363
FireRed-Image-Edit-1.1	firered_image_edit_1.1	4,806	2,406	2,400
LongCat-Image-Edit	longcat_image_edit	4,477	2,206	2,271
FLUX.2-dev	flux2_dev	4,480	2,207	2,273
Total		18,488	9,181	9,307

Split

In splits/split_seed2026.json (seed 2026), all 2,500 ImageNet image_id values are test. COCO supplies the train and validation image ids.

Files

<edit_model>/
  manifest.jsonl
  source/          JPEG, suffix .jpeg
  normal/          PNG
  abnormal/        PNG
  masks/normal/    PNG
  masks/abnormal/  PNG

Filenames follow {sample_id}__{language}_{role}, with roles source, normal, abnormal, normal_mask, and abnormal_mask. image_id is the ImageNet filename stem, such as n01440764_12971.

manifest.jsonl fields used to load a row:

Field	Meaning
sample_id	Source image, target object, and anomaly id
image_id	ImageNet stem, also the split key
language	zh or en
edit_model_name	Editor directory name
anomaly_type	One category below
iou	IoU between the normal and abnormal masks
normal_prompt, abnormal_prompt	Instructions applied to produce this row
normal_prompt_zh, normal_prompt_en, abnormal_prompt_zh, abnormal_prompt_en	Bilingual prompt fields stored with the row

source_paths and export_paths are absolute paths from the original export machine. Load images from source/, normal/, abnormal/, and masks/ in this folder.

Anomaly categories

Id	Pairs	Share	Abnormal edit
C1_physical_attribute	4,126	22.3%	Material or physical state is not one the object can have
C2_spatial_relation	2,542	13.7%	Support, contact, containment, or scale is inconsistent
C3_function_affordance	3,150	17.0%	The object does not support the use shown in the edit
C4_scene_context	2,618	14.2%	A recognizable object does not belong in the scene
C5_biological_structure	3,216	17.4%	Anatomy, part count, or bodily configuration is inconsistent
C6_interaction_relation	2,836	15.3%	An action or relation between entities is incoherent

The normal edit replaces the same object in a semantically ordinary way. Both edits are meant to keep location, scale, lighting, and viewpoint close to each other.

ImageNet photographs remain under the ImageNet terms. This folder does not relicense them.