State-of-the-art Image & Video CLIP, Multimodal Large Language Models, and More!

Hiera: A fast, powerful, and simple hierarchical vision transformer.

A deep learning library for video understanding research.

PySlowFast: video understanding codebase from FAIR for reproducing state-of-the-art video models.

Code release for "Spatiotemporal Residual Networks for Video Action Recognition", NIPS16 & "Spatiotemporal Multiplier Networks for Video Action Recognition", CVPR17.

Code release for "Convolutional Two-Stream Network Fusion for Video Action Recognition", CVPR 2016.

Sublinear memory optimization for deep learning, reduce GPU memory cost to train deeper nets

Code for Visualizing Deep Convolutional Neural Networks Using Natural Pre-images, IJCV 2016

Macro to draw architecture of convolutional neural nets in tikz.

Fully Convolutional Instance-aware Semantic Segmentation