AI Paper Atlas

Research map / Video / Video-language understanding

MoVA: Learning Asymmetric Dual Projections for Modular Long Video-Text Alignment

Peiyuan Zhu, Shaoan Xie, Zijian Li, Yifan Shen, Namrata Deka, Harsh Shrivastava, Guangyi Chen, Kun Zhang

ECCV 2026 · Video › Video-language understanding · approach: Video-language pretraining · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add temporal mask network and concept mask network (dual asymmetric projections) vs global loss only (ablation)CLIP-based video-text model · ActivityNet T2VR@1: 43.7 -> 47.6Improved
use dual projections instead of SmartCLIP-style single-direction visual-to-text masking (ablation)MoVA · ActivityNet T2VR@1: 35.6 -> 47.6Improved
apply temporal masks on the global text representation rather than raw tokens (ablation)MoVA · ActivityNet T2VR@1: 42.0 -> 47.6Improved
add DSL post-processingMoVA · ActivityNet T2VR@1: 47.8 -> 53.6Improved
remove different-frame same-mask loss (ablation)MoVA · ActivityNet T2VR@1: -3.6 R@1 when removedHurt
use soft masks instead of hard masks (ablation)MoVA · ActivityNet T2VR@1: 47.8 -> 47.4No clear effect
increase frames from 128 to 160 (ablation)MoVA · ActivityNet T2VR@1: 47.8 -> 46.6Hurt
increase TMN depth beyond a shallow depth (ablation)MoVA · ActivityNet T2VR@1: 47.8 (2 blocks) -> 47.0 (8 blocks)No clear effect
set global-to-modular loss ratio above 1.2 (ablation)MoVA · loss weighting sweepretrievalHurt
use MoVA text encoder in VideoCrafter2 instead of CLIPVideoCrafter2 · VBenchaverage score: 0.8456 -> 0.8708Improved
MoVA vs VideoCLIP-XL on short-caption DiDeMoMoVA · DiDeMo T2VR@1: 38.6 -> 57.5Improved

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