Deep Learning Models for Voice Emotion Analysis
Select a model to analyze speech emotions. Each model uses different techniques and has unique strengths.
Please select a model from the Models tab to start analyzing emotions.
This project was developed collaboratively by William Sander and Seth Shienbrood as part of CS 3540 at the University of Vermont. While we worked together throughout the project, sharing insights and helping refine each other's approaches, we each focused on complementary methodologies.
CRNN Spectrogram-Based Approach
Feature Extraction-Based Approach
Cao, H., Cooper, D. G., Keutmann, M. K., Gur, R. C., Nenkova, A., & Verma, R. (2014). CREMA-D: Crowd-sourced emotional multimodal actors dataset. IEEE Transactions on Affective Computing, 5(4), 377–390. GitHub
Burkhardt, F., Paeschke, A., Rolfes, M., Sendlmeier, W. F., & Weiss, B. (2005). A database of German emotional speech. Proceedings of Interspeech, 1517–1520.
Livingstone, S. R., & Russo, F. A. (2018). The Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS). PloS One, 13(5), e0196391. Zenodo
Jackson, P., & Haq, S. (2008). Surrey Audio-Visual Expressed Emotion (SAVEE) Database. University of Surrey, UK. Official site
Pichora-Fuller, M. K., Dupuis, K., & Reed, M. (2020). Toronto Emotional Speech Set (TESS). University of Toronto.
Busso, C., et al. (2008). IEMOCAP: Interactive emotional dyadic motion capture database. Language Resources and Evaluation, 42(4), 335–359. Access page
Graves, A., & Schmidhuber, J. (2005). Framewise phoneme classification with bidirectional LSTM networks. Neural Networks, 18(5–6), 602–610.
Bahdanau, D., Cho, K., & Bengio, Y. (2014). Neural machine translation by jointly learning to align and translate. arXiv:1409.0473.
Park, D. S., et al. (2019). SpecAugment: A simple data augmentation method for ASR. Interspeech, 2613–2617. arXiv
McFee, B., et al. (2015). librosa: Audio and music signal analysis in Python. Proc. 14th Python in Science Conference, 18–25. librosa.org
Paszke, A., et al. (2019). PyTorch: An imperative style, high-performance deep learning library. NeurIPS, 32. pytorch.org