cd/entity/Ref-Youtube-VOS· home› entities› Ref-Youtube-VOS
grep -l @ref-youtube-vos /news/*.json | wc -l → 1

Ref-Youtube-VOS

mentions 1 type Organization feed RSS

// recent coverage 1 mentions

04:00
2026-09-25
arxiv.org
computer-vision

MoVISA: Multi-Token Reasoning for Video Object Segmentation

Researchers developed MoVISA, a multi-token reasoning method for video object segmentation that replaces the single SEG token used in Multimodal Large Language Model pipelines with multiple tokens suc…

// co-occurs with top 4 entities
MoVISA 1 MeViS 1 DAVIS17 1 ReVOS 1
// topics top 5 topics
computer vision 1 artificial intelligence 1 machine learning 1 large language models 1 ai research 1