AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Test-Time Retrieval-Augmented Adaptation for Vision-Language Models

Xinqi Fan, Xueli Chen, Luoxiao Yang, Chuin Hong Yap, Rizwan Qureshi, Qi Dou, Moi Hoon Yap, Mubarak Shah

ICCV 2025 · Vision-language and multimodal › CLIP and zero-shot · approach: Test-time adaptation of VLMs · 9 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 SMGD module to the TDA one-shot baseline (ablation)TDA baseline · Cross-domain benchmark, ViT-B/16top-1 accuracy: 67.04 -> 68.46Improved
add SMGD module to the TDA one-shot baseline (ablation)TDA baseline · Cross-domain benchmark, ResNet-50top-1 accuracy: 59.82 -> 61.03Improved
add MRA module to the TDA one-shot baseline (ablation)TDA baseline · Cross-domain benchmark, ViT-B/16top-1 accuracy: 67.04 -> 67.49Improved
add MRA module to the TDA one-shot baseline (ablation)TDA baseline · Cross-domain benchmark, ResNet-50top-1 accuracy: 59.82 -> 60.02Improved
combine SMGD and MRA over the baseline (ablation)TDA baseline · Cross-domain benchmark, ViT-B/16top-1 accuracy: 67.04 -> 68.76Improved
combine SMGD and MRA over the baseline (ablation)TDA baseline · Cross-domain benchmark, ResNet-50top-1 accuracy: 59.82 -> 61.25Improved
use a nonzero update coefficient eta (0.65) instead of eta = 0 (ablation)TT-RAA · UCF101top-1 accuracy: eta = 0 gives poor adaptation; best eta = 0.65Improved
raise update coefficient eta beyond 0.65 (ablation)TT-RAA · UCF101top-1 accuracy: performance deteriorates beyond 0.65Hurt
build the Gaussian database from the target training set (TT-RAA-A) (ablation)TT-RAA · Cross-domain benchmark (Fig. 5b)top-1 accuracyImproved
use TT-RAA instead of TDATDA · EuroSATtop-1 accuracy: 58.00 -> 66.12Improved
use TT-RAA instead of TDATDA · UCF101top-1 accuracy: 70.66 -> 71.29Improved

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