Abstract:Prototypical Network is a popular few-shot solver that aims at establishing a feature metric generalizable to novel few-shot classification (FSC) tasks using deep neural networks. However, its performance drops dramatically when generalizing to the FSC tasks in new domains. In this study, we revisit this problem and argue that the devil lies in the simplicity bias pitfall in neural networks. In specific, the network tends to focus on some biased shortcut features (e.g., color, shape, etc.) that are exclusively sufficient to distinguish very few classes in the meta-training tasks within a pre-defined domain, but fail to generalize across domains as some desirable semantic features. To mitigate this problem, we propose a Local-global Distillation Prototypical Network (LDP-net). Different from the standard Prototypical Network, we establish a two-branch network to classify the query image and its random local crops, respectively. Then, knowledge distillation is conducted among these two branches to enforce their class affiliation consistency. The rationale behind is that since such global-local semantic relationship is expected to hold regardless of data domains, the local-global distillation is beneficial to exploit some cross-domain transferable semantic features for feature metric establishment. Moreover, such local-global semantic consistency is further enforced among different images of the same class to reduce the intra-class semantic variation of the resultant feature. In addition, we propose to update the local branch as Exponential Moving Average (EMA) over training episodes, which makes it possible to better distill cross-episode knowledge and further enhance the generalization performance. Experiments on eight cross-domain FSC benchmarks empirically clarify our argument and show the state-of-the-art results of LDP-net. Code is available in https://github.com/NWPUZhoufei/LDP-Net

Few-Shot Learning with Prototype Rectification for Cross-Domain Hyperspectral Image Classification

Cross-Domain Few-Shot Hyperspectral Image Classification With Cross-Modal Alignment and Supervised Contrastive Learning

Cross-Domain Few-Shot Learning Based on Graph Convolution Contrast for Hyperspectral Image Classification

Cross-Domain Self-Taught Network for Few-Shot Hyperspectral Image Classification

Few-shot learning based on prototype rectification with a self-attention mechanism

Cross-Domain Few-Shot Learning Based on Feature Disentanglement for Hyperspectral Image Classification

Cross-Domain Few-Shot Hyperspectral Image Classification with Class-Wise Attention

Cross-Domain Meta-Learning Under Dual-Adjustment Mode for Few-Shot Hyperspectral Image Classification

Adaptive Parametric Prototype Learning for Cross-Domain Few-Shot Classification

Adaptive Domain-Adversarial Few-Shot Learning for Cross-Domain Hyperspectral Image Classification

Tensor-Based Few-Shot Learning for Cross-Domain Hyperspectral Image Classification

Heterogeneous Few-Shot Learning for Hyperspectral Image Classification

Residual channel attention based sample adaptation few-shot learning for hyperspectral image classification

Reweighting and Information-Guidance Networks for Few-Shot Learning

Few-Shot Learning With Mutual Information Enhancement for Hyperspectral Image Classification

Boosting Few-Shot Hyperspectral Image Classification Using Pseudo-Label Learning

Graph-Based Domain Adaptation Few-Shot Learning for Hyperspectral Image Classification

ALPN: Active-Learning-Based Prototypical Network for Few-Shot Hyperspectral Imagery Classification

Revisiting Prototypical Network for Cross Domain Few-Shot Learning

Prototype Rectification with Region-Wise Foreground Enhancement for Few-Shot Classification.

Membership-Grade Based Prototype Rectification for Fine-Grained Few-Shot Classification