AI Paper Atlas

Research map / Video / Video segmentation and tracking

MeViS: A Multi-Modal Dataset for Referring Motion Expression Video Segmentation

Henghui Ding, Chang Liu, Shuting He, Kaining Ying, Xudong Jiang, Chen Change Loy, Yu–Gang Jiang

TPAMI 2025 · Video › Video segmentation and tracking · approach: Matching-based video object segmentation · 20 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 a large language model for motion reasoning (ablation)LMPM · MeViSv2J&F: 38.3 -> 42.4Improved
add a large language model for motion reasoning (ablation)LMPM · MeViSv1J&F: 42.2 -> 45.9Improved
add temporal-level contrastive loss (ablation)LMPM+LLM · MeViSv2J&F: 42.4 -> 43.9Improved
add temporal-level contrastive loss (ablation)LMPM+LLM · MeViSv1J&F: 45.9 -> 47.6Improved
add temporal context attention to image-based VLT (ablation)VLT · MeViSJ&F: 27.8 -> 35.5Improved
add temporal context attention to ReferFormer (ablation)ReferFormer · MeViSJ&F: 31.0 -> 36.3Improved
add temporal context to VLT (ablation)VLT · Refer-YouTube-VOSJ&F: 63.1 -> 62.7No clear effect
add temporal context to VLT (ablation)VLT · DAVIS17-RVOSJ&F: 60.4 -> 60.3No clear effect
train on referring image datasets instead of video dataVLT · MeViSJ&F: 27.8 -> 22.5Hurt
use adaptive multi-object output instead of top-1 maskDsHmp/LMPM++ · MeViSv2N-acc.: 0.0 -> 21.0Improved
add detection head for direct box predictionLMPM++ · MeViSv2 RMOTHOTA*: 38.1 -> 38.8Improved
use ground-truth text vs audio inputLMPM · MeViSv2 AVOSJ&F: 38.3 -> 37.5Mixed

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