Test-Time Adaptation of Quantized ViTs via Single-Pass Quantizer-Aligned Recalibration
Oct 6, 2026·
,,·
1 min read
Hyeongheon Cha
Young D. Kwon
Sung-Ju Lee
QuAR method overview (
).Position
First author
Project Goals & Works
- Improve the robustness of quantized vision transformers under distribution shift while keeping the model and quantizers frozen.
- Recalibrate activations at late-block quantizer inputs and the classifier input using running channel statistics and source calibration statistics.
- Adapt within a single forward pass, without backpropagation, parameter updates, or storage of past samples.
Key Results
- On ImageNet-C at severity 5, improved ViT-B/16 accuracy over the strongest evaluated backpropagation-free baseline by 4.00 percentage points at W3A3 and 2.28 points at W8A8.
- Used about 170 KB of state and reduced latency by at least 46% compared with the evaluated two-forward-pass methods.
- Evaluated robustness across continual streams, label shifts, additional distribution shifts, and multiple architectures.