---

OpenGait is a flexible and extensible gait analysis project provided by the Shiqi Yu Group and supported in part by WATRIX.AI. The corresponding paper has been accepted by CVPR2023 as a highlight paper. The extension paper has been accepted to TPAMI2025.

What's New

• [Sep 2025] BiggerGait has been accepted to NeurIPS2025🎉 and is available at here. Here are checkpoints.
• [Jun 2025] Scoliosis1K-Pose has been accepted to MICCAI2025🎉. Extends ScoNet by introducing pose annotations and clinical priors for interpretable scoliosis screening. Dataset is available on the project homepage.
• [Jun 2025] LidarGait++ has been accepted to CVPR2025🎉 and open-source in configs/lidargaitv2.
• [Jun 2025] The extension paper of OpenGait, further strengthened by the advancements of DeepGaitV2, SkeletonGait, and SkeletonGait++, has been accepted for publication in TPAMI🎉. We sincerely acknowledge the valuable contributions and continuous support from the OpenGait community.
• [Feb 2025] The diffusion-based DenoisingGait has been accepted to CVPR2025🎉 Congratulations to Dongyang! This is his SECOND paper!
• [Feb 2025] Chao successfully defended his Ph.D. thesis in Oct. 2024🎉🎉🎉 You can access the full text in Chao's Thesis in English or 樊超的学位论文（中文版）.
• [Dec 2024] The multimodal MultiGait++ has been accepted to AAAI2025🎉 Congratulations to Dongyang! This is his FIRST paper!
• [Jun 2024] The first large-scale gait-based scoliosis screening benchmark ScoNet is accepted to MICCAI2024🎉 Congratulations to Zirui! This is his FIRST paper! The code is released here, and you can refer to project homepage for details.
• [May 2024] The code of Large Vision Model based method BigGait is available at here. CCPG's checkpoints.
• [Apr 2024] Our team's latest checkpoints for projects such as DeepGaitv2, SkeletonGait, SkeletonGait++, and SwinGait will be released on Hugging Face. Additionally, previously released checkpoints will also be gradually made available on it.
• [Mar 2024] Chao gives a talk about 'Progress in Gait Recognition'. The video and slides are both available😊
• [Mar 2024] The code of SkeletonGait++ is released here, and you can refer to readme for details.
• [Mar 2024] BigGait has been accepted to CVPR2024🎉 Congratulations to Dingqiang! This is his FIRST paper!
• [Jan 2024] The code of transfomer-based SwinGait is available at here.

Our Works

• [NeurIPS'25] BiggerGait: Unlocking Gait Recognition with Layer-wise Representations from Large Vision Models Paper, and BiggerGait Code.
• [MICCAI'25] Pose as Clinical Prior: Learning Dual Representations for Scoliosis Screening. Paper and Scoliosis1K Dataset.
• [CVPR'25] LidarGait++: Learning Local Features and Size Awareness from LiDAR Point Clouds for 3D Gait Recognition. Paper and LidarGait++ Code
• [TPAMI'25] OpenGait: A Comprehensive Benchmark Study for Gait Recognition Towards Better Practicality. Paper. This extension includes a key update with in-depth insights into emerging trends and challenges of gait recognition in Sec. VII.
• [CVPR'25] On Denoising Walking Videos for Gait Recognition. Paper and DenoisingGait Code
• [Chao's Thesis] Gait Representation Learning and Recognition, Chinese Original and English Translation.
• [AAAI'25] Exploring More from Multiple Gait Modalities for Human Identification, Paper and MultiGait++ Code.
• [TBIOM'24] A Comprehensive Survey on Deep Gait Recognition: Algorithms, Datasets, and Challenges, Survey Paper.
• [MICCAI'24] Gait Patterns as Biomarkers: A Video-Based Approach for Classifying Scoliosis, Paper, Scoliosis1K Dataset, and ScoNet Code.
• [CVPR'24] BigGait: Learning Gait Representation You Want by Large Vision Models. Paper, and BigGait Code.
• [AAAI'24] SkeletonGait++: Gait Recognition Using Skeleton Maps. Paper, and SkeletonGait++ Code.
• [AAAI'24] Cross-Covariate Gait Recognition: A Benchmark. Paper, CCGR Dataset, and ParsingGait Code.
• [Arxiv'23] Exploring Deep Models for Practical Gait Recognition. Paper, DeepGaitV2 Code, and SwinGait Code.
• [TPAMI'23] Learning Gait Representation from Massive Unlabelled Walking Videos: A Benchmark, Paper, GaitLU-1M Dataset, and GaitSSB Code.
• [CVPR'23] LidarGait: Benchmarking 3D Gait Recognition with Point Clouds, Paper, SUSTech1K Dataset and LidarGait Code.
• [CVPR'23] OpenGait: Revisiting Gait Recognition Toward Better Practicality, Highlight Paper, and GaitBase Code.
• [ECCV'22] GaitEdge: Beyond Plain End-to-end Gait Recognition for Better Practicality, Paper, and GaitEdge Code.

A Real Gait Recognition System: All-in-One-Gait

The workflow of All-in-One-Gait involves the processes of pedestrian tracking, segmentation and recognition. See here for details.

Highlighted features

• Multiple Dataset supported: CASIA-B, OUMVLP, SUSTech1K, HID, GREW, Gait3D, CCPG, CASIA-E, and GaitLU-1M.
• Multiple Models Support: We reproduced several SOTA methods and reached the same or even better performance.
• DDP Support: The officially recommended Distributed Data Parallel (DDP) mode is used during both the training and testing phases.
• AMP Support: The Auto Mixed Precision (AMP) option is available.
• Nice log: We use tensorboard and logging to log everything, which looks pretty.

Getting Started

Quick Start (uv)

# Install dependencies
uv sync --extra torch

# Train
CUDA_VISIBLE_DEVICES=0,1 uv run python -m torch.distributed.launch --nproc_per_node=2 opengait/main.py --cfgs ./configs/baseline/baseline.yaml --phase train

# Test
CUDA_VISIBLE_DEVICES=0,1 uv run python -m torch.distributed.launch --nproc_per_node=2 opengait/main.py --cfgs ./configs/baseline/baseline.yaml --phase test

Note: The --nproc_per_node argument must exactly match the number of GPUs specified in CUDA_VISIBLE_DEVICES. For single-GPU evaluation, use CUDA_VISIBLE_DEVICES=0 and --nproc_per_node=1 with the DDP launcher.

Please see 0.get_started.md. We also provide the following tutorials for your reference:

• Prepare dataset
• Detailed configuration
• Customize model
• Advanced usages

Model Zoo

✨✨✨You can find all the checkpoint files at ✨✨✨!

The result list of appearance-based gait recognition is available here.

The result list of pose-based gait recognition is available here.

Authors:

• Chao Fan (樊超), 12131100@mail.sustech.edu.cn
• Chuanfu Shen (沈川福), 11950016@mail.sustech.edu.cn
• Junhao Liang (梁峻豪), 12132342@mail.sustech.edu.cn

Now OpenGait is mainly maintained by Dongyang Jin (金冬阳), 11911221@mail.sustech.edu.cn

Acknowledgement

• GLN: Saihui Hou (侯赛辉)

• GaitGL: Beibei Lin (林贝贝)

• GREW: GREW TEAM

• FastPoseGait Team: FastPoseGait Team

• Gait3D Team: Gait3D Team

Citation

@InProceedings{Fan_2023_CVPR,
    author    = {Fan, Chao and Liang, Junhao and Shen, Chuanfu and Hou, Saihui and Huang, Yongzhen and Yu, Shiqi},
    title     = {OpenGait: Revisiting Gait Recognition Towards Better Practicality},
    booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    month     = {June},
    year      = {2023},
    pages     = {9707-9716}
}

@ARTICLE{fan2025opengait,
  author={Fan, Chao and Hou, Saihui and Liang, Junhao and Shen, Chuanfu and Ma, Jingzhe and Jin, Dongyang and Huang, Yongzhen and Yu, Shiqi},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence}, 
  title={OpenGait: A Comprehensive Benchmark Study for Gait Recognition Towards Better Practicality}, 
  year={2025},
  volume={},
  number={},
  pages={1-18},
  doi={10.1109/TPAMI.2025.3576283}
}

Note: This code is only used for academic purposes, people cannot use this code for anything that might be considered commercial use.