Abstract:Cross-domain hyperspectral image (HSI) classification can improve the model's classification performance in the target domain by utilizing the rich knowledge from the source domain. However, existing cross-domain HSI classification methods mostly belong to transductive learning, which is difficult to apply to domain generalization tasks where the target domain is unseen during model learning. Inspired by human causal reasoning and knowledge induction mechanisms, this article develops an inductive learning-based framework for hyperspectral domain generalization: Causal meta-generalization. By simulating domain generalization scenarios, the framework helps the model induct domain-invariant causal meta-knowledge, thereby ensuring its strong generalization ability to unseen target domains. Specifically, we first propose a bottleneck variational auto-encoder (B-VAE) based on a forward–reverse information bottleneck, decoupling the domain distribution and class distribution of HSIs. By perturbing the domain distribution to generate virtual domains, we simulate potential domain distribution changes in the real world, providing a data basis for the induction of causal meta-knowledge. Second, in the process of simulating domain generalization scenarios, we establish a dual-layer optimization mechanism (DLOM) based on invariant-generalization risk minimization. In the inner layer optimization, by minimizing the model's invariant causal effect loss (ICEL) in the virtual source domains, we guide the model to learn domain-invariant causal meta-knowledge. In the outer optimization, by minimizing the model's generalization risk in the unseen virtual target domain, we enhance the applicability of causal meta-knowledge in domain generalization tasks. This proposed method has potential applications in remote sensing signal-processing tasks, such as the recognition of crop pests and diseases and the identification of minerals. The code can be accessed at https://github.com/wzr78998/CMG.

Deep discriminative causal domain generalization

Causality-inspired Latent Feature Augmentation for Single Domain Generalization

Knowledge Distillation-based Domain-invariant Representation Learning for Domain Generalization

Unbiased Semantic Representation Learning Based on Causal Disentanglement for Domain Generalization

Causality Inspired Representation Learning for Domain Generalization

A Causality-Aware Perspective on Domain Generalization via Domain Intervention

Learning Domain-Invariant Relationship with Instrumental Variable for Domain Generalization

Causality-based Dual-Contrastive Learning Framework for Domain Generalization

Inducing Causal Meta-Knowledge From Virtual Domain: Causal Meta-Generalization for Hyperspectral Domain Generalization

Joint Domain Alignment and Discriminative Feature Learning for Unsupervised Deep Domain Adaptation

DIGIC: Domain Generalizable Imitation Learning by Causal Discovery

Learning Causal Representation for Training Cross-Domain Pose Estimator Via Generative Interventions

A Causal Inspired Early-Branching Structure for Domain Generalization

Towards Counterfactual Fairness-aware Domain Generalization in Changing Environments

Cross-Domain Gated Learning for Domain Generalization

Domain Generalization via Causal Adjustment for Cross-Domain Sentiment Analysis

Idag: Invariant DAG Searching for Domain Generalization

Rethinking Domain Generalization: Discriminability and Generalizability

Deep causal representation learning for unsupervised domain adaptation

Category-Stitch Learning for Union Domain Generalization

Meta-causal Learning for Single Domain Generalization