Score: 0

Mobile-Efficient Speech Emotion Recognition Using DistilHuBERT: A Cross-Corpus Validation Study

Published: December 29, 2025 | arXiv ID: 2512.23435v1

By: Saifelden M. Ismail

Speech Emotion Recognition (SER) has significant potential for mobile applications, yet deployment remains constrained by the computational demands of state-of-the-art transformer architectures. This paper presents a mobile-efficient SER system based on DistilHuBERT, a distilled and 8-bit quantized transformer that achieves 92% parameter reduction compared to full-scale Wav2Vec 2.0 models while maintaining competitive accuracy. We conduct a rigorous 5-fold Leave-One-Session-Out (LOSO) cross-validation on the IEMOCAP dataset to ensure speaker independence, augmented with cross-corpus training on CREMA-D to enhance generalization. Cross-corpus training with CREMA-D yields a 1.2% improvement in Weighted Accuracy, a 1.4% gain in Macro F1-score, and a 32% reduction in cross-fold variance, with the Neutral class showing the most substantial benefit at 5.4% F1-score improvement. Our approach achieves an Unweighted Accuracy of 61.4% with a quantized model footprint of only 23 MB, representing approximately 91% of full-scale baseline performance. Cross-corpus evaluation on RAVDESS reveals that the theatrical nature of acted emotions causes predictions to cluster by arousal level rather than valence: happiness is systematically confused with anger due to acoustic saturation in high-energy expressions. Despite this theatricality effect reducing overall RAVDESS accuracy to 43.29%, the model maintains robust arousal detection with 97% recall for anger and 64% for sadness. These findings establish a Pareto-optimal tradeoff between model size and accuracy, enabling practical affect recognition on resource-constrained mobile devices.

Category
Computer Science:
Sound