English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

It is a well-known fact that the performance of deep learning modelsdeteriorates when they encounter a distribution shift at test time. Test-timeadaptation (TTA) algorithms have been proposed to adapt the model online whileinferring test data. However, existing research predominantly focuses onclassification tasks through the optimization of batch normalization layers orclassification heads, but this approach limits its applicability to variousmodel architectures like Transformers and makes it challenging to apply toother tasks, such as object detection. In this paper, we propose a novel onlineadaption approach for object detection in continually changing test domains,considering which part of the model to update, how to update it, and when toperform the update. By introducing architecture-agnostic and lightweightadaptor modules and only updating these while leaving the pre-trained backboneunchanged, we can rapidly adapt to new test domains in an efficient way andprevent catastrophic forgetting. Furthermore, we present a practical andstraightforward class-wise feature aligning method for object detection toresolve domain shifts. Additionally, we enhance efficiency by determining whenthe model is sufficiently adapted or when additional adaptation is needed dueto changes in the test distribution. Our approach surpasses baselines on widelyused benchmarks, achieving improvements of up to 4.9\%p and 7.9\%p in mAP forCOCO $\rightarrow$ COCO-corrupted and SHIFT, respectively, while maintainingabout 20 FPS or higher.

What, How, and When Should Object Detectors Update in Continually Changing Test Domains?