Multimedia Computing Group, Nanjing University

Pinned Loading

1. VideoMAE VideoMAE Public

   [NeurIPS 2022 Spotlight] VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training

   Python 1.8k 173

2. VideoChat3 VideoChat3 Public

   VideoChat3: a generalist video MLLM built for fine-grained motion, long-form reasoning, temporal grounding, and online proactive response.

   Python 350 11

3. MixFormer MixFormer Public

   [CVPR 2022 Oral & TPAMI 2024] MixFormer: End-to-End Tracking with Iterative Mixed Attention

   Python 542 74

4. OneStreamer OneStreamer Public

   OneStreamer: Unifying Perception, Memory, and Proactive Response in Streaming Video Interaction

   Python 229 4

5. MOTIP MOTIP Public

   [CVPR 2025] Multiple Object Tracking as ID Prediction

   Python 579 56

6. VideoJev VideoJev Public

   VideoJev: System One Model for Video Understanding and Interaction

   Python 249 10