Random Registers for Cross-Domain Few-Shot Learning

TL;DR

This paper introduces the use of random registers to perturb attention in Vision Transformers, significantly improving cross-domain few-shot learning performance by promoting flatter minima and better transferability.

Contribution

It reveals that replacing learnable prompts with random registers enhances ViT transferability in CDFSL and proposes a simple method to perturb attention maps for improved results.

Findings

Random registers improve target-domain performance in CDFSL.

Perturbing attention with random registers leads to flatter loss landscapes.

The approach achieves state-of-the-art results on four benchmarks.

Abstract

Cross-domain few-shot learning (CDFSL) aims to transfer knowledge from a data-sufficient source domain to data-scarce target domains. Although Vision Transformer (ViT) has shown superior capability in many vision tasks, its transferability against huge domain gaps in CDFSL is still under-explored. In this paper, we find an intriguing phenomenon: during the source-domain training, prompt tuning, as a common way to train ViT, could be harmful for the generalization of ViT in target domains, but setting them to random noises (i.e., random registers) could consistently improve target-domain performance. We then delve into this phenomenon for an interpretation. We find that learnable prompts capture domain information during the training on the source dataset, which views irrelevant visual patterns as vital cues for recognition. This can be viewed as a kind of overfitting and increases the sharpness of the loss landscapes. In contrast, random registers are essentially a novel way of perturbing attention for the sharpness-aware minimization, which helps the model find a flattened minimum in loss landscapes, increasing the transferability. Based on this phenomenon and interpretation, we further propose a simple but effective approach for CDFSL to enhance the perturbation on attention maps by adding random registers on the semantic regions of image tokens, improving the effectiveness and efficiency of random registers. Extensive experiments on four benchmarks validate our rationale and state-of-the-art performance. Codes and models are available at https://github.com/shuaiyi308/REAP.