KBlueLeaf / TTVidT
Other195M2038Updated Sep 30, 2026
Downloads
203
Likes
8
Last updated
Sep 30, 2026
License
apache-2.0
TT-VidT is a self-supervised video encoder built for motion. A DINOv3 ViT-B/16 processes every frame independently (the appearance path), while a compact Temporal Transfer pathway turns each frame into a few motion tokens that exchange information across time. This repository …
Tasks & library
- video classification
- transformers
Tags
- transformers
- ttvidt
- feature-extraction
- video
- self-supervised-learning
- motion
Technical specifications
Parameters195M
Librarytransformers
LanguagesNot available
Base modelfacebook/dinov3-vitb16-pretrain-lvd1689m
Licenseapache-2.0