Speech Emotion Recognition
Audio Signal Processing
Extracts MFCCs, Chroma, and Mel-Spectrogram features from speech audio with Librosa, then classifies the speaker's emotional state — happy, sad, angry, neutral — using Scikit-Learn classifiers, with feature-distribution plots and a confusion matrix for evaluation.
Stack
- Librosa
- Scikit-Learn
- MLPClassifier
- Random Forest