Co-salient object detection uses rank-consistent set reasoning approach
Rank-Consistent Set Reasoning for Co-Salient Object Detection
Computer Vision and Pattern Recognition
Summary
Finding common important objects in a group of images can be tricky because not all images show those objects clearly. The authors showed a way to look at all images as a set instead of a list, which helps spot objects that clearly appear across many images. They taught their system to ignore strange or different images in the group and focus on what is common. Their method combines data from different scales in the images and tests how stable the detected objects are across the group. This helps accurately highlight shared important objects without relying on language hints or outside tools.
What this means in practice
- •For computer vision engineers: Develop image analysis tools that locate common objects across multiple related images without using language or external detectors.
- •For video content editors: Improve automated tools to find consistently important objects in video frames that share similar scenes for efficient editing.
Authors
Yuan Xiang, Matteo Rossi, Yingzhou Chen
Abstract
Co-salient object detection (Co-SOD) requires a model to find foreground regions that are salient in individual images and supported by the image group. We present \emph{Rank-Consistent Set Reasoning} (RCSR), a supervised dense-prediction framework that models a group as an unordered set rather than as a sequence of images or a semantic label. The core idea is to rank how strongly each spatial region agrees with a small collection of learned group slots at every image scale, and to aggregate these ranks with a robust trimmed statistic. This suppresses accidental pairwise matches and prevents one atypical group member from dominating the shared representation. A set encoder builds group slots directly from multi-scale visual features, while a rank-consistency gate measures whether the ordering of candidate regions is stable across group members. The gated slots are decoded jointly with per-image features to produce co-saliency maps. The model contains no natural-language branch, no open-vocabulary detector, and no external segmentation model. We further introduce a group permutation objective and hard-distractor augmentation so that the model learns the properties of a set-level target rather than memorizing image order or isolated visual saliency. We formulate an evaluation protocol for CoCA, CoSal2015, and CoSOD3k, together with tests of group-size robustness, distractor rejection, order invariance, and cross-dataset transfer.