---
license: other
language:
  - en
  - zh
pretty_name: PAPE
size_categories:
  - 10K<n<100K
tags:
  - image-editing
  - image-forensics
  - forgery-localization
task_categories:
  - image-to-image
  - image-segmentation
---

# PAPE

PAPE (Plausible Anomalous Paired Edits) is a paired image-editing dataset for localizing semantic edits. Each sample edits one object in a real photograph twice: a semantically ordinary replacement, and a replacement that introduces one controlled semantic anomaly. Both edits have a binary change mask. Only pairs whose normal and abnormal masks have IoU of at least 0.5 are included.

Source photographs come from COCO, ImageNet, and Places. Those photographs stay under the terms of the source dataset. This repository does not relicense them. Generated edits and masks are released for research.

## Subsets

| Subset | Source images | Language-specific pairs | Role in `splits/split_seed2026.json` | Readme |
| --- | ---: | ---: | --- | --- |
| [COCO](COCO) | 2,000 | 14,668 | train 1,600 / val 200 / test 200 image ids | [COCO/README.md](COCO/README.md) |
| [ImageNet](ImageNet) | 2,500 | 18,488 | all 2,500 image ids are test | [ImageNet/README.md](ImageNet/README.md) |
| [Places](Places) | 2,000 | 14,591 | all 2,000 image ids are test | [Places/README.md](Places/README.md) |
| Total | 6,500 | 47,747 | seed 2026 |  |

A pair is one language. Chinese and English edits of the same object are separate rows, so one source image can contribute more than one pair, and not every image has both languages for every editor.

## Editors

Each subset contains the same four edit models:

| Directory | Model |
| --- | --- |
| `qwen_image_edit_2511` | Qwen-Image-Edit-2511 |
| `firered_image_edit_1.1` | FireRed-Image-Edit-1.1 |
| `longcat_image_edit` | LongCat-Image-Edit |
| `flux2_dev` | FLUX.2-dev |

## Layout

```text
<COCO|ImageNet|Places>/<edit_model>/
  manifest.jsonl
  source/
  normal/
  abnormal/
  masks/normal/
  masks/abnormal/
```

`manifest.jsonl` has one JSON object per line. `normal_prompt` and `abnormal_prompt` are the instructions used for that row. `image_id` joins a row to `splits/split_seed2026.json`. `source_paths` and `export_paths` record paths on the machine that built the release; use the files in this repository instead.

Details, counts, and anomaly categories are in each subset readme.