Cross-Domain Few-Shot Segmentation via Multi-view Progressive Adaptation

2026-02-05

AI Summary

This work addresses the performance bottleneck in cross-domain few-shot segmentation caused by scarce and insufficiently diverse target-domain samples, as well as substantial domain discrepancies. To tackle these challenges, we propose a multi-view progressive adaptation method that integrates a hybrid progressive augmentation strategy with a dual-path multi-view prediction architecture. Strong supervision is applied along the sequential path, while consistency constraints are enforced in the parallel path, thereby establishing an easy-to-hard learning curriculum. By generating diverse and complex views and jointly optimizing both data augmentation and learning strategies, our approach significantly enhances model robustness and accuracy on the target domain. Extensive experiments demonstrate that the proposed method achieves an average performance gain of 7.0% over state-of-the-art approaches across multiple cross-domain few-shot segmentation benchmarks.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Segmentation
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Web applications in cross-disciplinary domains and verticals such as mixed reality, smart cities, and digital health

Abstract

Cross-Domain Few-Shot Segmentation aims to segment categories in data-scarce domains conditioned on a few exemplars. Typical methods first establish few-shot capability in a large-scale source domain and then adapt it to target domains. However, due to the limited quantity and diversity of target samples, existing methods still exhibit constrained performance. Moreover, the source-trained model's initially weak few-shot capability in target domains, coupled with substantial domain gaps, severely hinders the effective utilization of target samples and further impedes adaptation. To this end, we propose Multi-view Progressive Adaptation, which progressively adapts few-shot capability to target domains from both data and strategy perspectives. (i) From the data perspective, we introduce Hybrid Progressive Augmentation, which progressively generates more diverse and complex views through cumulative strong augmentations, thereby creating increasingly challenging learning scenarios. (ii) From the strategy perspective, we design Dual-chain Multi-view Prediction, which fully leverages these progressively complex views through sequential and parallel learning paths under extensive supervision. By jointly enforcing prediction consistency across diverse and complex views, MPA achieves both robust and accurate adaptation to target domains. Extensive experiments demonstrate that MPA effectively adapts few-shot capability to target domains, outperforming state-of-the-art methods by a large margin (+7.0%).

Problem

• Cross-Domain Few-Shot Segmentation
• Domain Adaptation
• Few-Shot Learning
• Semantic Segmentation
• Data Scarcity

Innovation

• Cross-Domain Few-Shot Segmentation
• Progressive Adaptation
• Multi-view Learning
• Hybrid Augmentation
• Prediction Consistency