AI Paper Atlas

Research map / Video / Video-language understanding

UniViT: Unifying Image and Video Understanding in One Vision Encoder

Feilong Tang, Xiang An, Haolin Yang, Yin Xie, Kaicheng Yang, Ming Hu, Zheng Cheng, Xingyu Zhou, Zimin Ran, Imran Razzak, Ziyong Feng, Behzad Bozorgtabar and 2 more

NeurIPS 2025 · Video › Video-language understanding · approach: Long video understanding · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes UniViT, a cluster-driven unified self-supervised vision encoder intended to capture both spatial details and temporal dynamics for image and video understanding.

Problem

Existing vision pretraining methods are biased toward either spatial modeling (e.g., CLIP) or temporal modeling (e.g., V-JEPA), limiting joint capture of spatial details and temporal dynamics.

Approach

A cluster-driven unified self-supervised learning framework that captures structured semantics across images and videos within a single vision encoder (abstract is truncated, so details beyond this are unknown).

Experimental setup

Baselines: CLIP V-JEPA

Builds on: CLIP V-JEPA

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy