Re-thinking Co-Salient Object Detection
Deng-Ping Fan, Tengpeng Li, Zheng Lin, Ge-Peng Ji, Dingwen Zhang, Ming-Ming Cheng, Huazhu Fu, Jianbing Shen
TL;DR
CoSOD tasks suffer from data bias toward appearance similarity in existing datasets. The paper introduces CoSOD3k, a large-scale, semantically diverse dataset with 3,316 images in 160 groups and rich hierarchical annotations, and CoEG-Net, a unified CoSOD framework that uses a PCA-based co-attention projection alongside EGNet-style saliency priors. It also provides a broad benchmark of 40 algorithms across iCoSeg, CoSal2015, and CoSOD3k, highlighting that traditional SOD models frequently compete with or surpass CoSOD baselines on some datasets, while CoEG-Net achieves state-of-the-art performance on the challenging CoSOD3k. The work delivers a standard benchmark, a trainable CoSOD baseline, and detailed analyses of scalability, stability, and metric design, offering a foundation for future end-to-end CoSOD research. Overall, the dataset and framework advance understanding of cross-image co-saliency and facilitate more robust, semantically aware detection in real-world scenarios.
Abstract
In this paper, we conduct a comprehensive study on the co-salient object detection (CoSOD) problem for images. CoSOD is an emerging and rapidly growing extension of salient object detection (SOD), which aims to detect the co-occurring salient objects in a group of images. However, existing CoSOD datasets often have a serious data bias, assuming that each group of images contains salient objects of similar visual appearances. This bias can lead to the ideal settings and effectiveness of models trained on existing datasets, being impaired in real-life situations, where similarities are usually semantic or conceptual. To tackle this issue, we first introduce a new benchmark, called CoSOD3k in the wild, which requires a large amount of semantic context, making it more challenging than existing CoSOD datasets. Our CoSOD3k consists of 3,316 high-quality, elaborately selected images divided into 160 groups with hierarchical annotations. The images span a wide range of categories, shapes, object sizes, and backgrounds. Second, we integrate the existing SOD techniques to build a unified, trainable CoSOD framework, which is long overdue in this field. Specifically, we propose a novel CoEG-Net that augments our prior model EGNet with a co-attention projection strategy to enable fast common information learning. CoEG-Net fully leverages previous large-scale SOD datasets and significantly improves the model scalability and stability. Third, we comprehensively summarize 40 cutting-edge algorithms, benchmarking 18 of them over three challenging CoSOD datasets (iCoSeg, CoSal2015, and our CoSOD3k), and reporting more detailed (i.e., group-level) performance analysis. Finally, we discuss the challenges and future works of CoSOD. We hope that our study will give a strong boost to growth in the CoSOD community. The benchmark toolbox and results are available on our project page at http://dpfan.net/CoSOD3K/.