On-Device Realistic Test-Time Adaptation via Bias-Resistant Statistical Alignment
Haojie Bai, Aiguo Chen, Ruiting Dai, Yijia Rong, Zirui Wang, Jiaxin Liu, Kexin Li, Schahram Dustdar
Abstract
Test-Time Adaptation (TTA) aims to adapt pretrained models to unseen test data, which is crucial for resource-constrained edge devices that must handle distribution shifts on the fly without human supervision. However, conventional TTA methods often fail in realistic scenarios characterized by continuous shifts and severe class imbalances while incurring prohibitive memory overheads. To enable continuous adaptation to test data on edge devices under a realistic TTA setting, we propose a Bias-Resistant Online Statistical Alignment (BOSA) method. BOSA facilitates unbiased adaptation via a discrepancy-aware statistical alignment mechanism integrated with a class-balanced memory bank. To ensure memory efficiency, we further design a saliency-guided activation sparsification and gradient reconstruction scheme, which drastically reduces memory overhead without sacrificing gradient integrity. Extensive evaluations demonstrate that BOSA achieves superior accuracy with a compact memory consumption compared to state-of-the-art methods under realistic TTA settings.
BibTeX
@inproceedings{ijcai2026_ondevicerealisti,
title = {On-Device Realistic Test-Time Adaptation via Bias-Resistant Statistical Alignment},
author = {Haojie Bai and Aiguo Chen and Ruiting Dai and Yijia Rong and Zirui Wang and Jiaxin Liu and Kexin Li and Schahram Dustdar},
booktitle = {IJCAI 2026},
year = {2026}
}