Sense-X/UniFormer
A unified transformer model for efficient spatiotemporal visual representation learning across classification, detection, segmentation, and pose estimation.
Not currently ranked — collecting fresh signals.
star history
UniFormer is the official implementation of papers published at ICLR2022 and TPAMI2023, proposing a unified architecture that bridges local convolution and global self-attention for visual recognition. It provides pretrained models and training code for multiple vision tasks including image classification, video classification, object detection, semantic segmentation, and pose estimation, with both standard and lightweight model variants.
Frequently asked
- What is Sense-X/UniFormer?
- A unified transformer model for efficient spatiotemporal visual representation learning across classification, detection, segmentation, and pose estimation.
- Is UniFormer open source?
- Yes — Sense-X/UniFormer is open source, released under the Apache-2.0 license.
- What language is UniFormer written in?
- Sense-X/UniFormer is primarily written in Python.
- How popular is UniFormer?
- Sense-X/UniFormer has 906 stars on GitHub.
- Where can I find UniFormer?
- Sense-X/UniFormer is on GitHub at https://github.com/Sense-X/UniFormer.