Preprocessing of audio files for sentiment analysis.
Models used for classification of audios:
- Neural Network
- SVM: sklearn.svm.SVC
- Random Forest Classifier
The accuracy of the models is close to 100%. However, when a real audio file is tested the results are disappointing. A larger database with real and natural emotions should be used in order to be able to get accurate predictions under all circumstances.
- David Andres
This project is licensed under the MIT License - see the LICENSE file for details
- Reconocimiento de Estados Emocionales de Personas Mediante la Voz Utilizando Algoritmos de Aprendizaje de Máquina
- Database: Toronto emotional speech set (TESS)
- Confusion matrix for NN approach: scikit-learn documentation example