Pinned Loading

1. AutoGaze automatically removes redundant patches in a video, reducing #tokens in ViT/MLLM by 4x-100x.

   Python 301 24

2. Scaling Vision Pre-Training to 4K Resolution

   Python 226 12

3. When do we not need larger vision models?

   Python 419 16

4. VILA is a family of state-of-the-art vision language models (VLMs) for diverse multimodal AI tasks across the edge, data center, and cloud.

   Python 3.9k 330

5. Official code for "TOAST: Transfer Learning via Attention Steering"

   Python 187 10

6. Official code for "Top-Down Visual Attention from Analysis by Synthesis" (CVPR 2023 highlight)

   Jupyter Notebook 165 14