ToneSense

The project leverages Librosa for audio feature extraction, including MFCCs and spectral features and uses TensorFlow to develop a deep learning model for classifying emotions like happiness, sadness and anger. It incorporates spectrogram analysis for visual insights and supports real-time emotion recognition for practical applications.

Execution Guide:

1. Clone the repository:

   git clone https://github.com/kr1shnasomani/ToneSense.git
   cd ToneSense
   

2. Download the dependencies:

   pip install -r requirements
   

3. Run the code and it will save the model with the name model.keras

Accuracy & Loss Over Epochs:

Model Prediction:

Input (the file is in .mp4 format as GitHub doesn't support audio files):

OAF_back_fear.mp4

Output: