Performance Optimization and Interpretability Analysis of Lightweight Transformer Models for Speech Emotion Recognition Using Open-Source Emotional Speech Datasets
Academic Journal of Science and Technology · 13.03.2026 · 19 (3) · 104-112
DOI: 10.54097/1y0ms617 ↗Kurz gefasst
Fachartikel in der Zeitschrift Academic Journal of Science and Technology, online erschienen am 13.03.2026. Verfasst von Shaoyang Zhang. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Accurate speech-emotion recognition (SER) remains challenging across speakers, languages, and recording conditions, especially under tight computing conditions. We present a lightweight three-class SER framework that unifies acted corpora and pairs frozen self-supervised speech embeddings with a compact residual CNN. A Multi-Corpus Unified Labelling Protocol (MCULP) harmonises CREMA-D, RAVDESS, and EmoDB into a balanced three-class taxonomy (negative, sad_fear, pos_neutral), yielding 7,618 utterances with an 80/10/10 speaker-independent split. Our 3.9-M-parameter Residual Squeeze-and-Excitation 1-D CNN (ResAttn1D-CNN) uses five residual blocks with channel attention and 768×400 wav2vec 2.0-base embeddings. A Tri-Aug pipeline—Gaussian noise, random crop-pad, and SpecAugment-style temporal masking—improves robustness. Trained with AdamW and mixed precision, the model converges in ≈10 hours. On the held-out test set, it reaches 62.9% accuracy and 0.627 macro-F1, outperforming a strong three-layer CNN by 20.9 points and exceeding prior CNN results on comparable tasks. Ablations: +4.9 pp (attention), +2.9 pp (Tri-Aug), +3.2 pp (depth); the confusion matrix shows residual ambiguity between low-arousal negative and neutral speech. Real-time inference (
Zitieren
APA
Zhang, S. (2026). Performance Optimization and Interpretability Analysis of Lightweight Transformer Models for Speech Emotion Recognition Using Open-Source Emotional Speech Datasets. Academic Journal of Science and Technology, 19(3), 104-112. https://doi.org/10.54097/1y0ms617
Vancouver
Zhang S. Performance Optimization and Interpretability Analysis of Lightweight Transformer Models for Speech Emotion Recognition Using Open-Source Emotional Speech Datasets. Academic Journal of Science and Technology. 2026;19(3):104-112. doi:10.54097/1y0ms617
BibTeX
@article{Zhang2026,
author = {Zhang, Shaoyang},
title = {Performance Optimization and Interpretability Analysis of Lightweight Transformer Models for Speech Emotion Recognition Using Open-Source Emotional Speech Datasets},
journal = {Academic Journal of Science and Technology},
year = {2026},
volume = {19},
number = {3},
pages = {104--112},
doi = {10.54097/1y0ms617}
}
EUVIMED