EAGLE: Eigen Aggregation Learning for Object-Centric Unsupervised Semantic Segmentation

Chanyoung Kim, Woojung Han, Dayun Ju, Seong Jae Hwang

Keywords: Object-centric, Unsupervised Semantic Segmentation

Abstract: Semantic segmentation has innately relied on extensive pixel-level annotated data, leading to the emergence of unsupervised methodologies. Among them, leveraging self-supervised Vision Transformers for unsupervised semantic segmentation (USS) has been making steady progress with expressive deep features. Yet, for semantically segmenting images with complex objects, a predominant challenge remains: the lack of explicit object-level semantic encoding in patch-level features particularly with diverse structures. To address this gap, we present a novel approach, which emphasizes object-centric representation learning for unsupervised semantic segmentation, named EAGLE.

Submission Number: 5