Title:Adaptive Hardness-driven Augmentation and Alignment Strategies for Multi-Source Domain Adaptations

Abstract:Multi-source Domain Adaptation (MDA) aims to transfer knowledge from multiple labeled source domains to an unlabeled target domain. Nevertheless, traditional methods primarily focus on achieving inter-domain alignment through sample-level constraints, such as Maximum Mean Discrepancy (MMD), neglecting three pivotal aspects: 1) the potential of data augmentation, 2) the significance of intra-domain alignment, and 3) the design of cluster-level constraints. In this paper, we introduce a novel hardness-driven strategy for MDA tasks, named "A3MDA" , which collectively considers these three aspects through Adaptive hardness quantification and utilization in both data Augmentation and domain this http URL achieve this, "A3MDA" progressively proposes three Adaptive Hardness Measurements (AHM), i.e., Basic, Smooth, and Comparative AHMs, each incorporating distinct mechanisms for diverse scenarios. Specifically, Basic AHM aims to gauge the instantaneous hardness for each source/target sample. Then, hardness values measured by Smooth AHM will adaptively adjust the intensity level of strong data augmentation to maintain compatibility with the model's generalization this http URL contrast, Comparative AHM is designed to facilitate cluster-level constraints. By leveraging hardness values as sample-specific weights, the traditional MMD is enhanced into a weighted-clustered variant, strengthening the robustness and precision of inter-domain alignment. As for the often-neglected intra-domain alignment, we adaptively construct a pseudo-contrastive matrix by selecting harder samples based on the hardness rankings, enhancing the quality of pseudo-labels, and shaping a well-clustered target feature space. Experiments on multiple MDA benchmarks show that " A3MDA " outperforms other methods.