Weakly-supervised Contrastive Learning for Unsupervised Object Discovery

TL;DR

This paper introduces a novel weakly-supervised contrastive learning approach to improve unsupervised object discovery by enhancing semantic feature extraction and using PCA for object localization, outperforming existing methods.

Contribution

It proposes a semantic-guided self-supervised learning model based on WCL and PCA for effective unsupervised object discovery, addressing limitations of previous generative and clustering methods.

Findings

Effective object localization using PCA and eigenvalues.

Enhanced semantic feature extraction via WCL and fine-tuning DINO.

Superior performance on benchmark datasets.

Abstract

Unsupervised object discovery (UOD) refers to the task of discriminating the whole region of objects from the background within a scene without relying on labeled datasets, which benefits the task of bounding-box-level localization and pixel-level segmentation. This task is promising due to its ability to discover objects in a generic manner. We roughly categorise existing techniques into two main directions, namely the generative solutions based on image resynthesis, and the clustering methods based on self-supervised models. We have observed that the former heavily relies on the quality of image reconstruction, while the latter shows limitations in effectively modeling semantic correlations. To directly target at object discovery, we focus on the latter approach and propose a novel solution by incorporating weakly-supervised contrastive learning (WCL) to enhance semantic information exploration. We design a semantic-guided self-supervised learning model to extract high-level semantic features from images, which is achieved by fine-tuning the feature encoder of a self-supervised model, namely DINO, via WCL. Subsequently, we introduce Principal Component Analysis (PCA) to localize object regions. The principal projection direction, corresponding to the maximal eigenvalue, serves as an indicator of the object region(s). Extensive experiments on benchmark unsupervised object discovery datasets demonstrate the effectiveness of our proposed solution. The source code and experimental results are publicly available via our project page at https://github.com/npucvr/WSCUOD.git.