My internship assignment involves creating a research documentation,prototype documentation and the Python script that takes audio files as input for Aggression detection in speech or audio.

This Aggression Detection Prototype is designed to detect aggression in audio data using machine learning. It combines audio-based feature extraction with emotion classification to identify aggression in spoken content. The prototype utilizes a Multi-Layer Perceptron (MLP) classifier and is designed to provide users with a simple tool for emotion detection.

Audio-based feature extraction including MFCC, chroma, and mel features. Training an MLP classifier for emotion classification. User-friendly interface for audio input and emotion prediction. Accuracy and F1 score reporting. Model serialization for future use.

Before using the prototype, make sure you have the following prerequisites:

Python 3.7 or higher installed. Required Python libraries: librosa, soundfile, numpy, scikit-learn, pandas.

Audio Files: Place your audio files in the project directory. Supported formats include MP3, WAV, and other common audio formats containing spoken content. Analysis Feature Extraction: The prototype extracts audio features (MFCC, chroma, mel) from the provided audio files.

Model Training: The MLP classifier is trained on the extracted features to classify emotions, with a focus on detecting aggression.

Accuracy: The prototype reports an accuracy score to indicate how well the model performs in aggression detection.

F1 Scores: F1 scores for individual emotions (e.g., 'calm,' 'disgust,' 'happy,' 'fearful') are provided.

Emotion Predictions: Users can predict emotions for their audio data using the trained model. A sample prediction ('disgust') is demonstrated in the code.

Model Serialization: The trained model is saved to a file for future use.

For detailed documentation of the prototype, including its approach, features, preprocessing steps, and user manual, please refer to the Prototype Documentation file.

This project is licensed under the GNU License - see the LICENSE file for details.