lucidrains/TimeSformer-pytorch
A PyTorch implementation of TimeSformer, an attention-based deep learning model for video classification.
Not currently ranked — collecting fresh signals.
star history
This repository implements TimeSformer from Facebook AI, a pure attention-based architecture for video understanding. It uses divided space-time attention, applying attention along the time axis before spatial attention. The model processes video tensors of shape (batch x frames x channels x height x width) and outputs classification predictions. It is built with PyTorch and supports variable-length video batching via masking.
Frequently asked
- What is lucidrains/TimeSformer-pytorch?
- A PyTorch implementation of TimeSformer, an attention-based deep learning model for video classification.
- Is TimeSformer-pytorch open source?
- Yes — lucidrains/TimeSformer-pytorch is open source, released under the MIT license.
- What language is TimeSformer-pytorch written in?
- lucidrains/TimeSformer-pytorch is primarily written in Python.
- How popular is TimeSformer-pytorch?
- lucidrains/TimeSformer-pytorch has 730 stars on GitHub.
- Where can I find TimeSformer-pytorch?
- lucidrains/TimeSformer-pytorch is on GitHub at https://github.com/lucidrains/TimeSformer-pytorch.