10 papers
Autoencoder based optimized SSL representations: Complexity Minimization and improved Dysarthric ASR
Paban Sapkota, Hemant Kumar Kathania, Mikko Kurimo +2
Self-supervised learning (SSL) models extract rich speech representations but often come with high-dimensional features, increasing computational complexity. This work explores an…
Cross-Dataset, Age, and Gender Generalization: A Comprehensive Analysis of Fine-Tuning Strategies for Low-Resource Children's ASR
Abhijit Sinha, Hemant Kumar Kathania, Sudarsana Reddy Kadiri +1
The challenge associated with recognizing dysarthric speech primarily arises from pronounced acoustic variability attributed to impaired articulatory precision. Past research has d…
DSSCNet: A Transfer Learning Framework for Cross-Corpus Dysarthric Speech Severity Classification
Arnab Kumar Roy, Hemant Kumar Kathania, Paban Sapkota +2
Dysarthric speech severity classification is challenging due to speaker variability, class imbalance, and limited datasets. This study introduces DSSCNet, a deep learning model tha…
How Well Do Self-Supervised Speech Models Encode Age and Gender in Children's Speech? A Layer-Wise Analysis Across Multiple Architectures
Abhijit Sinha, Hemant Kumar Kathania, Mohit Joshi +3
Self-supervised learning (SSL) models have become a central component of modern speech processing systems, as they enable the learning of rich acoustic representations without reli…
Improving End-to-End Speech Recognition for Dysarthric Speech through In-Domain Data Augmentation
Paban Sapkota, Hemant Kumar Kathania, Sudarsana Reddy Kadiri +1
Dysarthric speech recognition is crucial for facilitating effective communication among individuals with dysarthria. However, accurately recognizing dysarthric speech poses signifi…
Systematic Study of Dysarthric Speech Recognition: Spectral Features and Acoustic Models
Paban Sapkota, Hemant Kumar Kathania, Mikko Kurimo +2
The challenge associated with recognizing dysarthric speech primarily arises from pronounced acoustic variability attributed to impaired articulatory precision. Past research has d…