Unsupervised Part Discovery via Descriptor-Based Masked Image Restoration with Optimized Constraints

TL;DR

This paper introduces Masked Part Autoencoder (MPAE), an unsupervised method that effectively discovers meaningful object parts across diverse categories and scenarios by restoring masked patches with learned descriptors, overcoming previous robustness limitations.

Contribution

The paper proposes MPAE, a novel unsupervised framework that learns part descriptors and restores masked image regions, enabling robust part discovery across multiple categories and complex scenarios.

Findings

Robustly discovers meaningful parts across various categories.

Effective in handling occlusion and complex scenarios.

Achieves superior performance compared to existing methods.

Abstract

Part-level features are crucial for image understanding, but few studies focus on them because of the lack of fine-grained labels. Although unsupervised part discovery can eliminate the reliance on labels, most of them cannot maintain robustness across various categories and scenarios, which restricts their application range. To overcome this limitation, we present a more effective paradigm for unsupervised part discovery, named Masked Part Autoencoder (MPAE). It first learns part descriptors as well as a feature map from the inputs and produces patch features from a masked version of the original images. Then, the masked regions are filled with the learned part descriptors based on the similarity between the local features and descriptors. By restoring these masked patches using the part descriptors, they become better aligned with their part shapes, guided by appearance features from unmasked patches. Finally, MPAE robustly discovers meaningful parts that closely match the actual object shapes, even in complex scenarios. Moreover, several looser yet more effective constraints are proposed to enable MPAE to identify the presence of parts across various scenarios and categories in an unsupervised manner. This provides the foundation for addressing challenges posed by occlusion and for exploring part similarity across multiple categories. Extensive experiments demonstrate that our method robustly discovers meaningful parts across various categories and scenarios. The code is available at the project https://github.com/Jiahao-UTS/MPAE.