Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Subspace Optimization for Backpropagation-Free Continual Test-Time Adaptation

About

We introduce PACE, a backpropagation-free continual test-time adaptation system that directly optimizes the affine parameters of normalization layers. Existing derivative-free approaches struggle to balance runtime efficiency with learning capacity, as they either restrict updates to input prompts or require continuous, resource-intensive adaptation regardless of domain stability. To address these limitations, PACE leverages the Covariance Matrix Adaptation Evolution Strategy with the Fastfood projection to optimize high-dimensional affine parameters within a low-dimensional subspace, leading to superior adaptive performance. Furthermore, we enhance the runtime efficiency by incorporating an adaptation stopping criterion and a domain-specialized vector bank to eliminate redundant computation. Our framework achieves state-of-the-art accuracy across multiple benchmarks under continual distribution shifts, reducing runtime by over 50% compared to existing backpropagation-free methods.

Damian S\'ojka, Sebastian Cygert, Marc Masana• 2026

Related benchmarks

TaskDatasetResultRank
Image ClassificationImageNet-R
Accuracy64.5
217
Image ClassificationImageNet-C
Accuracy67
117
Image ClassificationAverage IN-C, IN-R, DN-126
Average Accuracy62.5
10
Continual Test-Time AdaptationImageNet-C long-term continual adaptation--
10
Showing 4 of 4 rows

Other info

Follow for update