Style-Content Metric Learning for Multidomain Remote Sensing Object Recognition
Wenda Zhao, Ruikai Yang, Yu Liu, You He
Abstract
Previous remote sensing recognition approaches predominantly perform well on the training-testing dataset. However, due to large style discrepancies not only among multidomain datasets but also within a single domain, they suffer from obvious performance degradation when applied to unseen domains. In this paper, we propose a style-content metric learning framework to address the generalizable remote sensing object recognition issue. Specifically, we firstly design an inter-class dispersion metric to encourage the model to make decision based on content rather than the style, which is achieved by dispersing predictions generated from the contents of both positive sample and negative sample and the style of input image. Secondly, we propose an intra-class compactness metric to force the model to be less style-biased by compacting classifier's predictions from the content of input image and the styles of positive sample and negative sample. Lastly, we design an intraclass interaction metric to improve model's recognition accuracy by pulling in classifier's predictions obtained from the input image and positive sample. Extensive experiments on four datasets show that our style-content metric learning achieves superior generalization performance against the state-of-the-art competitors. Code and model are available at: https://github.com/wdzhao123/TSCM.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
Builds on6
- Deep Domain-Adversarial Image Generation for Domain GeneralisationKaiyang Zhou, Yongxin Yang, Timothy M. Hospedales, Tao XiangAAAI 2020 · 488 citations
- Learning to Diversify for Single Domain GeneralizationZijian Wang, Yadan Luo, Ruihong Qiu, Zi Huang et al.ICCV 2021 · 339 citations
- Maximum-Entropy Adversarial Data Augmentation for Improved Generalization and RobustnessLong Zhao, Ting Liu, Xi Peng, Dimitris N. MetaxasNeurIPS 2020 · 207 citations
- Reducing Domain Gap by Reducing Style BiasHyeonseob Nam, HyunJae Lee, Jongchan Park, Wonjun Yoon et al.CVPR 2021
- On Feature Normalization and Data AugmentationBoyi Li, Felix Wu, Ser-Nam Lim, Serge J. Belongie et al.CVPR 2021
Related papers
- En-Compactness: Self-Distillation Embedding & Contrastive Generation for Generalized Zero-Shot LearningXia Kong, Zuodong Gao, Xiaofan Li, Ming Hong et al.CVPR 2022 · 70 citations
- Style Projected Clustering for Domain Generalized Semantic SegmentationWei Huang, Chang Chen, Yong Li, Jiacheng Li et al.CVPR 2023
- ASGS: Single-Domain Generalizable Open-Set Object Detection via Adaptive Subgraph SearchingYuxuan Yuan, Luyao Tang, Yixin Chen, Chaoqi Chen et al.ICCV 2025 · 1 citation
- Deep Multi-Level Contrastive Clustering for Multi-Modal Remote Sensing ImagesWeiqi Liu, Yongshan Zhang, Xinxin Wang, Lefei ZhangACM MM 2025 · 2 citations
- Spatial-Temporal Context Model for Remote Sensing Imagery CompressionJinxiao Zhang, Runmin Dong, Juepeng Zheng, Mengxuan Chen et al.ACM MM 2024 · 6 citations
