Int*-Match: Balancing Intra-Class Compactness and Inter-Class Discrepancy for Semi-Supervised Speaker Recognition
Xingmei Wang, Jinghan Liu, Jiaxiang Meng, Boquan Li, Zijian Liu
Abstract
Open-set speaker recognition is to identify whether the voices are from the same speaker. One challenge of speaker recognition is collecting large amounts of high-quality data. Based on the promising results of image classification, one intuitively feasible solution is semi-supervised learning (SSL) which uses confidence thresholds to assign pseudo labels for unlabeled data. However, we empirically demonstrated that applying SSL methods to speaker recognition is non-trivial. These methods focus solely on inter-class discrepancy as thresholds to select pseudo labels, overlooking intra-class compactness, which is particularly important for open-set speaker recognition tasks. Motivated by this, we propose Int*-Match, a semi-supervised speaker recognition method selecting reliable pseudo labels with intra-class compactness and inter-class discrepancy for speaker recognition. In particular, we use the inter-class discrepancy of labeled data as the threshold for pseudo-label selection and adjust the threshold based on the intra-class compactness of the pseudo labels dynamically and adaptively. Our systematic experiments demonstrate the superiority of Int*-Match, presenting an outstanding Equal Error Rate (EER) of 1.00% on the VoxCeleb1 original test set, which is merely 0.06% below the performance achieved by fully supervised learning.
BibTeX
@article{Wang_Liu_Meng_Li_Liu_2025, title={Int*-Match: Balancing Intra-Class Compactness and Inter-Class Discrepancy for Semi-Supervised Speaker Recognition}, volume={39}, url={https://ojs.aaai.org/index.php/AAAI/article/view/34729}, DOI={10.1609/aaai.v39i24.34729}, abstractNote={Open-set speaker recognition is to identify whether the voices are from the same speaker. One challenge of speaker recognition is collecting large amounts of high-quality data. Based on the promising results of image classification, one intuitively feasible solution is semi-supervised learning (SSL) which uses confidence thresholds to assign pseudo labels for unlabeled data. However, we empirically demonstrated that applying SSL methods to speaker recognition is non-trivial. These methods focus solely on inter-class discrepancy as thresholds to select pseudo labels, overlooking intra-class compactness, which is particularly important for open-set speaker recognition tasks. Motivated by this, we propose Int*-Match, a semi-supervised speaker recognition method selecting reliable pseudo labels with intra-class compactness and inter-class discrepancy for speaker recognition. In particular, we use the inter-class discrepancy of labeled data as the threshold for pseudo-label selection and adjust the threshold based on the intra-class compactness of the pseudo labels dynamically and adaptively. Our systematic experiments demonstrate the superiority of Int*-Match, presenting an outstanding Equal Error Rate (EER) of 1.00% on the VoxCeleb1 original test set, which is merely 0.06% below the performance achieved by fully supervised learning.}, number={24}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Wang, Xingmei and Liu, Jinghan and Meng, Jiaxiang and Li, Boquan and Liu, Zijian}, year={2025}, month={Apr.}, pages={25407-25415} }