HSE: Hybrid Species Embedding for Deep Metric Learning

Yang, Bailin; Sun, Haoqiang; Li, Frederick W. B.; Chen, Zheng; Cai, Jianlu; Song, Chao

Abstract

Deep metric learning is crucial for finding an embedding function that can generalize to training and testing data, including unknown test classes. However, limited training samples restrict the model's generalization to downstream tasks. While adding new training samples is a promising solution, determining their labels remains a significant challenge. Here, we introduce Hybrid Species Embedding (HSE), which employs mixed sample data augmentations to generate hybrid species and provide additional training signals. We demonstrate that HSE outperforms multiple state-of-the-art methods in improving the metric Recall@K on the CUB-200 , CAR-196 and SOP datasets, thus offering a novel solution to deep metric learning's limitations.

Citation

Yang, B., Sun, H., Li, F. W. B., Chen, Z., Cai, J., & Song, C. (in press). HSE: Hybrid Species Embedding for Deep Metric Learning.

• Conference Name: International Conference on Computer Vision
• Conference Location: Paris
• Start Date: Oct 2, 2023
• End Date: Oct 6, 2023
• Acceptance Date: Aug 11, 2023
• Deposit Date: Sep 12, 2023
• Publicly Available Date: Sep 27, 2023
• Public URL: https://durham-repository.worktribe.com/output/1735635