This repository redistributes a PREPROCESSED version (HDF5 packed body crops and face crops) of the third-party LTCC person re-identification dataset, which is released by its original authors under a research-only license/agreement. By requesting access you confirm that (1) you have already obtained the original LTCC dataset from its official source and agreed to its terms, (2) you will use these files for non-commercial academic research only, (3) you will not redistribute them, and (4) you will cite the original dataset and our paper in any publication.

LTCC H5: preprocessed LTCC with body and face crops

Preprocessed HDF5 pack of the LTCC long-term cloth-changing person re-identification benchmark, as used in our paper FusionAgent: A Multimodal Agent with Dynamic Model Selection for Human Recognition (arXiv:2603.26908, code).

File	Size	Content
`ltcc.h5`	2.2 GB	body images, variable size (e.g. `(419,278,3)`)
`ltcc_face.h5`	0.4 GB	face crops, `(112,112,3)`

LTCC has no separate split files: the three top-level H5 groups train / query / test are the official splits, and each image name encodes <pid>_<clothes>_c<cam>_<frame>.

Download

pip install -U huggingface_hub
hf auth login   # you must have been granted access to this gated repo
hf download Paipile/ltcc-h5 --repo-type dataset --local-dir ./data/LTCC_ReID

Point an LTCC loader at ./data/LTCC_ReID; enumerate images with h5[split].keys() instead of listing the train/, query/, test/ folders.

H5 format

Keys mirror the directory structure of the original dataset; <img> is the original image file name without extension. Every leaf is an HDF5 dataset holding one RGB uint8 array of shape (H, W, 3) (i.e. np.asarray(PIL.Image.open(...).convert('RGB'))). Face crops are aligned 112×112 crops detected inside each body image; face_<k> enumerates multiple detections in the same image (most images have a single face_0). Images without a detectable face are simply absent from the face file.

Open the files with h5py.File(path, 'r') and keep the handle open across __getitem__ calls (one handle per worker when using multiprocess DataLoader workers). The arrays are stored uncompressed for fast random access.

File	Key path	Shape
`ltcc.h5`	`{train,query,test}/<img>`	`(H,W,3)`
`ltcc_face.h5`	`{train,query,test}/<img>/face_<k>`	`(112,112,3)`

import h5py
from PIL import Image
body = h5py.File('data/LTCC_ReID/ltcc.h5', 'r'); face = h5py.File('data/LTCC_ReID/ltcc_face.h5', 'r')
names = sorted(body['query'].keys())                       # e.g. '001_1_c11_015833'
img = Image.fromarray(body['query'][names[0]][:])
if names[0] in face['query']:
    faces = [Image.fromarray(face['query'][names[0]][k][:]) for k in face['query'][names[0]]]

License and terms

This repository contains derived data only; all rights to the images remain with the original authors. LTCC (Qian et al., Long-Term Cloth-Changing Person Re-identification, ACCV 2020) is research-only and must be requested from the authors: https://naiq.github.io/LTCC_Perosn_ReID.html. You must obtain the original dataset and accept its terms before using these files. Access here is gated: non-commercial academic research only, no redistribution.

Citation

If you use these files, please cite the original dataset above and our paper:

@article{zhu2026fusionagent,
  title   = {FusionAgent: A Multimodal Agent with Dynamic Model Selection for Human Recognition},
  author  = {Zhu, Jie and Guo, Xiao and Su, Yiyang and Jain, Anil and Liu, Xiaoming},
  journal = {arXiv preprint arXiv:2603.26908},
  year    = {2026}
}

Downloads last month: 27

Total file size:

2.56 GB

Paper for Paipile/ltcc-h5

FusionAgent: A Multimodal Agent with Dynamic Model Selection for Human Recognition

Paper • 2603.26908 • Published Mar 27