Rethinking preventing class-collapsing in metric learning with margin-based losses

• Main Authors: Levi, Elad, Xiao, Tete, Wang, Xiaolong, Darrell, Trevor
• Format: Conference Proceeding
• Language: English
• Subjects: Computer vision; Image and video retrieval; Image retrieval; Noise measurement; Representation learning; Training; Visualization
• ISSN: 2380-7504
• DOI: 10.1109/ICCV48922.2021.01015

Metric learning seeks perceptual embeddings where visually similar instances are close and dissimilar instances are apart, but learned representations can be sub-optimal when the distribution of intra-class samples is diverse and distinct sub-clusters are present. Although theoretically with optimal assumptions, margin-based losses such as the triplet loss and margin loss have a diverse family of solutions. We theoretically prove and empirically show that under reasonable noise assumptions, margin-based losses tend to project all samples of a class with various modes onto a single point in the embedding space, resulting in class collapse that usually renders the space ill-sorted for classification or retrieval. To address this problem, we propose a simple modification to the embedding losses such that each sample selects its nearest same-class counterpart in a batch as the positive element in the tuple. This allows for the presence of multiple sub-clusters within each class. The adaptation can be integrated into a wide range of metric learning losses. Our method demonstrates clear benefits on various fine-grained image retrieval datasets over a variety of existing losses; qualitative retrieval results show that samples with similar visual patterns are indeed closer in the embedding space.