AI Paper Atlas

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

RMAdapter: Reconstruction-based Multi-Modal Adapter for Vision-Language Models

Xiang Lin, Weixin Li, Shu Guo, Lihong Wang, Di Huang

AAAI 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Test-time adaptation of VLMs · 0 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 reconstruction branch and consistency constraint to MMA (ablation)MMA · Base-to-Novel, avg of 11 datasetsHM: 79.87 -> 80.62Improved
add consistency constraint alone to MMA (ablation)MMA · Base-to-Novel, avg of 11 datasetsHM: base 83.20 -> 83.86; novel 76.80 -> 76.31Mixed
add text reconstruction on top of constraints (ablation)MMA + constraints · Base-to-Novel, avg of 11 datasetsHM: 79.91 -> 80.51Improved
add visual reconstruction on top of constraints (ablation)MMA + constraints · Base-to-Novel, avg of 11 datasetsHM: 79.91 -> 80.39Improved
share down-projection instead of independent projections (ablation)RMAdapter · Base-to-Novel, avg of 11 datasetsHM: 80.52 -> 80.62Improved
share down-projection instead of sharing up-projection (ablation)RMAdapter · Base-to-Novel, avg of 11 datasetsHM: 80.30 -> 80.62Improved
use two-layer instead of single-layer up-projection (ablation)RMAdapter · Base-to-Novel, avg of 11 datasetsHM: 80.57 -> 80.62Improved
use three-layer instead of two-layer up-projection (ablation)RMAdapter · Base-to-Novel, avg of 11 datasetsHM: 80.62 -> 80.08Hurt
use RMAdapter instead of MMARMAdapter · Cross-dataset (ImageNet-trained, 10 targets zero-shot)avg accuracy: 66.61 -> 67.56Improved
use RMAdapter instead of CoPromptRMAdapter · Base-to-Novel HM, EuroSATHM: 85.84 vs 84.69Hurt
use RMAdapter instead of CoPromptRMAdapter · Domain generalization (ImageNet variants avg)avg accuracy: 60.42 -> 60.71Improved

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