# AG-ReID2023 — Data

来源：https://www.kaggle.com/competitions/ag-reid2023/data

采集时间：2026-10-19T07:40:15Z

**Important:** 
Before downloading, please read **Privacy and Usage Policy for AG-ReID dataset** at [https://www.kaggle.com/competitions/ag-reid2023/overview/privacy-policy](https://www.kaggle.com/competitions/ag-reid2023/overview/privacy-policy)

# Data
## Train
There will be 51,530 training images in JPG format depicting 807 identities.

## Test
The test set comprises two subsets:
1. `exp3_aerial_to_ground`: 808 ids | 4,348 query images | 19,259 gallery images. 
2. `exp6_ground_to_aerial`: 808 ids | 4,151  query images | 21,214 gallery images. 

## Annotations
The following table shows the annotation details for each image, including image name, person ID, camera, frame, and notes.
|Image Name | Person ID | Camera | Frame | Note |
| --- | --- |
| P0001T03220A1C3F01111.jpg |  P0001T03220A1 | C3 | F01111| C0: aerial, C2: wearable, C3: CCTV |

## Files
- `bounding_box_train` - contains JPG images of people for training the model.
- `exp3_aerial_to_ground` and `exp6_ground_to_aerial` contain query and bounding_box_test for evaluating person reID model
- `qut_attribute_v8.mat` - contains 88 binary attributes derived from 15 soft-biometric attributes for all 1615 identities. Use if required.
- `sample_submission.csv`  - contains the submission template. img_id from row 2 to row 4349 are queries of the exp3_aerial_to_ground set, and img_id from row 4350 to row 8500 are queries of the exp6_ground_to_aerial set.
- `test.csv` - contains test template for submission.
- `ReID2023_authorkit` - contains the template for the method description document.