PartCo: Part-Level Correspondence Priors Enhance Category Discovery
This work addresses the challenge of identifying both known and novel categories in unlabeled data for computer vision applications, representing an incremental improvement by integrating part-level cues into existing GCD methods.
The paper tackles the problem of Generalized Category Discovery (GCD) by introducing PartCo, a framework that incorporates part-level visual feature correspondences to enhance category discovery, achieving state-of-the-art results on multiple benchmark datasets.
Generalized Category Discovery (GCD) aims to identify both known and novel categories within unlabeled data by leveraging a set of labeled examples from known categories. Existing GCD methods primarily depend on semantic labels and global image representations, often overlooking the detailed part-level cues that are crucial for distinguishing closely related categories. In this paper, we introduce PartCo, short for Part-Level Correspondence Prior, a novel framework that enhances category discovery by incorporating part-level visual feature correspondences. By leveraging part-level relationships, PartCo captures finer-grained semantic structures, enabling a more nuanced understanding of category relationships. Importantly, PartCo seamlessly integrates with existing GCD methods without requiring significant modifications. Our extensive experiments on multiple benchmark datasets demonstrate that PartCo significantly improves the performance of current GCD approaches, achieving state-of-the-art results by bridging the gap between semantic labels and part-level visual compositions, thereby setting new benchmarks for GCD. Project page: https://visual-ai.github.io/partco