Domain Adaptation from Generated Multi-Weather Images for Unsupervised Maritime Object Classification

2025-01-26

AI Summary

To address challenges in unsupervised maritime target classification—including long-tailed distributions of object categories and weather conditions, severe label scarcity, and significant source-to-target domain shift—this paper proposes a domain adaptation framework tailored for real-world oceanic scenarios. Methodologically, it introduces two novel datasets: the multi-weather synthetic dataset AIMO and the real-world long-tailed dataset RMO. It is the first work to integrate vision-language models (specifically CLIP) into unsupervised domain adaptation, thereby enhancing cross-domain generalization for rare classes and under adverse weather. The approach synergistically combines large-model-based image generation, long-tailed learning, and vision-language alignment mechanisms. Experimental results demonstrate that while maintaining overall classification accuracy, the method significantly improves recognition performance for few-shot targets under extreme weather conditions. Both code and datasets are fully open-sourced.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Adversarial Attacks & Robustness
• Application Domains: Mobility, Driving & Flight

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

The classification and recognition of maritime objects are crucial for enhancing maritime safety, monitoring, and intelligent sea environment prediction. However, existing unsupervised methods for maritime object classification often struggle with the long-tail data distributions in both object categories and weather conditions. In this paper, we construct a dataset named AIMO produced by large-scale generative models with diverse weather conditions and balanced object categories, and collect a dataset named RMO with real-world images where long-tail issue exists. We propose a novel domain adaptation approach that leverages AIMO (source domain) to address the problem of limited labeled data, unbalanced distribution and domain shift in RMO (target domain), and enhance the generalization of source features with the Vision-Language Models such as CLIP. Experimental results shows that the proposed method significantly improves the classification accuracy, particularly for samples within rare object categories and weather conditions. Datasets and codes will be publicly available at https://github.com/honoria0204/AIMO.

Problem

• Marine Object Recognition
• Variable Weather Conditions
• Image Classification

Innovation

• Data Augmentation
• CLIP Tool
• Weather-Resilient Recognition