JDAI-CV/fast-reid
A research platform implementing state-of-the-art re-identification algorithms for person re-ID, image retrieval, and face recognition tasks.
Not currently ranked — collecting fresh signals.
star history
FastReID is a PyTorch-based computer vision research toolbox that implements state-of-the-art re-identification methods. It provides training pipelines, evaluation tools, and pretrained models for matching people or objects across different camera views. The platform supports various backbones including CNNs and Vision Transformers, and includes features like model distillation, AMP training, and hyperparameter optimization for computer vision research.
Frequently asked
- What is JDAI-CV/fast-reid?
- A research platform implementing state-of-the-art re-identification algorithms for person re-ID, image retrieval, and face recognition tasks.
- Is fast-reid open source?
- Yes — JDAI-CV/fast-reid is open source, released under the Apache-2.0 license.
- What language is fast-reid written in?
- JDAI-CV/fast-reid is primarily written in Python.
- How popular is fast-reid?
- JDAI-CV/fast-reid has 4k stars on GitHub.
- Where can I find fast-reid?
- JDAI-CV/fast-reid is on GitHub at https://github.com/JDAI-CV/fast-reid.