AI Paper Atlas

Research map / Video / Video segmentation and tracking

Live Interactive Training for Video Segmentation

Xinyu Yang, Haozheng Yu, Yihong Sun, Bharath Hariharan, Jennifer J. Sun

CVPR 2026 · Video › Video segmentation and tracking · approach: Matching-based video object segmentation · 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 LIT-LoRA online adaptationSAM2 · VOSTcorrections per video: 27.43 -> 18.24Improved
add LIT-LoRA online adaptationSAM2 · VOSTannotation time (min): 18.42 -> 12.91Improved
add LIT-LoRA at stricter threshold tau_IoU=0.75SAM2 · datasetscorrections: 17-24% reductionImproved
add LIT to DAM4SAMDAM4SAM · VOSTcorrections: 34.60 -> 22.46Improved
add LIT to SAMURAISAMURAI · VOSTcorrections: 26.96 -> 21.23Improved
add LIT-LoRA to CLIP online classificationCLIP ViT-B/32 · Stanford Carscorrections per class: 13.38 -> 7.57Improved
fine-tune original decoder at each correction instead of separate LoRA (ablation)LIT · VOSTcorrections per video: 27.43 -> 32.47Hurt
continual LoRA learning across corrections (vs first correction only) (ablation)LIT-LoRA · VOSTcorrections per video: 21.43 -> 18.24Improved
full decoder fine-tuning instead of LoRA (ablation)LIT · VOSTcorrections per video: 18.24 -> 17.46Improved
maintain three LoRAs instead of one (ablation)LIT-LoRA · VOSTcorrections per video: 18.24 -> 17.87No clear effect
increase training epochs from 40 to 60-100 (ablation)LIT-LoRA · VOSTcorrection reduction: < 2 pp gainNo clear effect
add LIT-LoRA in user study with 6 volunteersSAM2 demo · VOST user studycorrection number: 41.92% fewer correctionsImproved

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