Unsupervised Salient Object Detection with Spectral Cluster Voting
This addresses the problem of detecting salient objects without labeled data for computer vision applications, but it is incremental as it builds on existing self-supervised models and spectral clustering techniques.
The paper tackles unsupervised salient object detection by using spectral clustering on self-supervised features and a voting mechanism to select masks, resulting in a method called SelfMask that outperforms prior approaches on three benchmarks.
In this paper, we tackle the challenging task of unsupervised salient object detection (SOD) by leveraging spectral clustering on self-supervised features. We make the following contributions: (i) We revisit spectral clustering and demonstrate its potential to group the pixels of salient objects; (ii) Given mask proposals from multiple applications of spectral clustering on image features computed from various self-supervised models, e.g., MoCov2, SwAV, DINO, we propose a simple but effective winner-takes-all voting mechanism for selecting the salient masks, leveraging object priors based on framing and distinctiveness; (iii) Using the selected object segmentation as pseudo groundtruth masks, we train a salient object detector, dubbed SelfMask, which outperforms prior approaches on three unsupervised SOD benchmarks. Code is publicly available at https://github.com/NoelShin/selfmask.