How rushil07pawar/speech-emotion-recogntion-using-mfcc works
This notebook-based Speech Emotion Recognition project takes English or Hindi audio datasets through feature extraction and model training, then offers emotion prediction through a Gradio interface. The README documents dataset downloading, augmentation, CNN/BiLSTM/SVM modeling, saved models, and interface capabilities; no notebook source was available to verify finer-grained calls or the exact wiring among those models.
Main entry points: Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb Model Training
Components 13
Data and Features
- English Datasets
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb - Hindi Dataset
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb - Audio Augmentation
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb - Audio Features
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb
Model Training
- Model Training
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb - Trained Models
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb
Prediction Interface
- Gradio Interface
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb - Audio Input
- Emotion Prediction
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb - Waveform and Probabilities
Speech_Emotion_Recognition_Using_Mfcc_SVM_and_CNN.ipynb
Other components
- Researcher
- KaggleHub
- Google Drive
Connections 12
- Researcher → Gradio Interface: uses
- KaggleHub → English Datasets: downloads
- Google Drive → Hindi Dataset: stores
- English Datasets → Model Training: provides data
- Hindi Dataset → Model Training: provides data
- Model Training → Audio Augmentation: applies augmentation
- Model Training → Audio Features: extracts features
- Model Training → Trained Models: saves models
- Audio Input → Gradio Interface: uploaded or recorded
- Gradio Interface → Emotion Prediction: requests prediction
- Gradio Interface → Trained Models: selects split
- Gradio Interface → Waveform and Probabilities: displays