Abstract
Cache-based test-time adaptation (TTA) for vision-language models is often hindered by background bias in global representations and unreliable entropy-based cache admission under representation variations. To address these limitations, we propose RADC, which enhances prototype learning through reliable dual caching. RADC introduces a Semantic Foreground Cache that aggregates category-consistent spatial evidence from CLIP representations, yielding foreground prototypes that complement the global cache while mitigating background interference. To reliably manage both caches, Gaussian Risk Admission models multi-view representations as diagonal Gaussian distributions and jointly considers class separation and feature uncertainty to prioritize reliable cache candidates. RADC integrates zero-shot logits with complementary global- and foreground-cache predictions for robust inference. Extensive experiments on cross-domain and out-of-distribution benchmarks demonstrate consistent state-of-the-art performance.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Huang, S., Chen, Y., Li, X., & Zhou, J. (2026). RADC: Risk-Aware Dual Caching for Vision-Language Test-Time Adaptation. https://omanscience.com/en/articles/radc-risk-aware-dual-caching-for-vision-language-test-time-adaptation
MLA 9
Huang, Siyu, et al. "RADC: Risk-Aware Dual Caching for Vision-Language Test-Time Adaptation." https://omanscience.com/en/articles/radc-risk-aware-dual-caching-for-vision-language-test-time-adaptation.
Chicago (author–date)
Huang, Siyu, Yueyong Chen, Xuejiao Li, and Jun Zhou. 2026. "RADC: Risk-Aware Dual Caching for Vision-Language Test-Time Adaptation." https://omanscience.com/en/articles/radc-risk-aware-dual-caching-for-vision-language-test-time-adaptation.
Harvard
Huang, S., Chen, Y., Li, X. and Zhou, J. (2026) 'RADC: Risk-Aware Dual Caching for Vision-Language Test-Time Adaptation', Available at: https://omanscience.com/en/articles/radc-risk-aware-dual-caching-for-vision-language-test-time-adaptation.
Vancouver
Huang S, Chen Y, Li X, Zhou J. RADC: Risk-Aware Dual Caching for Vision-Language Test-Time Adaptation. https://omanscience.com/en/articles/radc-risk-aware-dual-caching-for-vision-language-test-time-adaptation
IEEE
S. Huang, Y. Chen, X. Li, and J. Zhou, "RADC: Risk-Aware Dual Caching for Vision-Language Test-Time Adaptation," https://omanscience.com/en/articles/radc-risk-aware-dual-caching-for-vision-language-test-time-adaptation.