AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

VisCoP: Visual Probing for Video Domain Adaptation of Vision Language Models

Dominick Reilly, Manish Govind, Le Xue, Srijan Das

ECCV 2026 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Geometric alignment and synthetic-to-real adaptation · 2 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 visual probes to a frozen vision encoder (VisCoP) instead of finetuning VEVLM (VideoLLaMA3) · egocentric targetdelta target vs base: +0.57 (full finetune) -> +3.53Improved
finetune connector, vision encoder and LLM (ablation)VLM · egocentric sourcedelta source: -0.50Hurt
finetune vision encoder with connector for depth adaptation (ablation)VLM · depth sourcedelta source: -3.87Hurt
add VisCoP probes for depth adaptationVLM · depth targetdelta target: +16.44 (connector only) -> +19.27Improved
train vision encoder in robot control adaptation (ablation)VLM · robot control sourcedelta source: -8.87 (VE+LLM) vs -4.58 (VisCoP)Hurt
add VisCoP in robot control simulationVLM · VIMA-Benchavg success rate: 65.13 -> 67.82Improved
add VisCoP in real-world robot with xArm-Det joint trainingVLM · xArm real-worldavg success rate: 80.00 -> 96.67Improved
restrict probe interaction to spatial only (ablation)VisCoP · egocentricdelta target: +3.53 -> +1.83Hurt
restrict probe interaction to temporal only (ablation)VisCoP · egocentricdelta target: +3.53 -> -0.30Hurt
use KL regularization toward base VLM on depth adaptation (ablation)connector+VE finetuning · depthsource avg: 68.9 -> 73.0Improved
add source replay (up to 25%) on depth adaptation (ablation)connector+VE finetuning · depthsource avg: 68.9 -> 70.0Mixed
increase number of probes beyond 16 (ablation)VisCoP · three domainsdelta target / sourceNo clear effect

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