PALM: Pushing Adaptive Learning Rate Mechanisms for Continual Test-Time Adaptation

arXiv:2403.10650v1 Announce Type: cross
Abstract: Real-world vision models in dynamic environments face rapid shifts in domain distributions, leading to decreased recognition performance. Continual test-time adaptation (CTTA) directly adjusts a pre-trained source discriminative model to these changing domains using test data. A highly effective CTTA method involves applying layer-wise adaptive learning rates, and selectively adapting pre-trained layers. However, it suffers from the poor estimation of domain shift and the inaccuracies arising from the pseudo-labels. In this work, we aim to overcome …

abstract arxiv continual cs.cv cs.lg data domain domains dynamic environments face layer palm performance rate recognition test type vision vision models wise world