Layer-Wise Analysis of Self-Supervised Representations for Age and Gender Classification in Children's Speech
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912537396641792 |
|---|---|
| author | Sinha, Abhijit Kumar, Harishankar Joshi, Mohit Kathania, Hemant Kumar Narayanan, Shrikanth Kadiri, Sudarsana Reddy |
| author_facet | Sinha, Abhijit Kumar, Harishankar Joshi, Mohit Kathania, Hemant Kumar Narayanan, Shrikanth Kadiri, Sudarsana Reddy |
| contents | Children's speech presents challenges for age and gender classification due to high variability in pitch, articulation, and developmental traits. While self-supervised learning (SSL) models perform well on adult speech tasks, their ability to encode speaker traits in children remains underexplored. This paper presents a detailed layer-wise analysis of four Wav2Vec2 variants using the PFSTAR and CMU Kids datasets. Results show that early layers (1-7) capture speaker-specific cues more effectively than deeper layers, which increasingly focus on linguistic information. Applying PCA further improves classification, reducing redundancy and highlighting the most informative components. The Wav2Vec2-large-lv60 model achieves 97.14% (age) and 98.20% (gender) on CMU Kids; base-100h and large-lv60 models reach 86.05% and 95.00% on PFSTAR. These results reveal how speaker traits are structured across SSL model depth and support more targeted, adaptive strategies for child-aware speech interfaces. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_10332 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Layer-Wise Analysis of Self-Supervised Representations for Age and Gender Classification in Children's Speech Sinha, Abhijit Kumar, Harishankar Joshi, Mohit Kathania, Hemant Kumar Narayanan, Shrikanth Kadiri, Sudarsana Reddy Audio and Speech Processing Artificial Intelligence Human-Computer Interaction Machine Learning Sound Children's speech presents challenges for age and gender classification due to high variability in pitch, articulation, and developmental traits. While self-supervised learning (SSL) models perform well on adult speech tasks, their ability to encode speaker traits in children remains underexplored. This paper presents a detailed layer-wise analysis of four Wav2Vec2 variants using the PFSTAR and CMU Kids datasets. Results show that early layers (1-7) capture speaker-specific cues more effectively than deeper layers, which increasingly focus on linguistic information. Applying PCA further improves classification, reducing redundancy and highlighting the most informative components. The Wav2Vec2-large-lv60 model achieves 97.14% (age) and 98.20% (gender) on CMU Kids; base-100h and large-lv60 models reach 86.05% and 95.00% on PFSTAR. These results reveal how speaker traits are structured across SSL model depth and support more targeted, adaptive strategies for child-aware speech interfaces. |
| title | Layer-Wise Analysis of Self-Supervised Representations for Age and Gender Classification in Children's Speech |
| topic | Audio and Speech Processing Artificial Intelligence Human-Computer Interaction Machine Learning Sound |
| url | https://arxiv.org/abs/2508.10332 |