Emotion Detection - CA Project

Introduction

This project integrates speech emotion recognition, text detection, and a user-friendly voice-based UI to analyze emotions from audio recordings. It utilizes OpenAI Whisper for speech-to-text, Wav2Vec2 for emotion detection, and Tkinter for an interactive user interface.

Folder Structure

CA_PROJECT/
│── emorecognition/
│   │── m4atestfolder/  # Folder containing test audio files
│   │   │── stressed.m4a
│   │   │── suprise.m4a
│   │   │── test.m4a
│   │   │── test2.m4a
│   │   │── test4.m4a
│   │   │── test5.m4a
│   │── __init__.py
│   │── emreco.py  # Main emotion recognition script
│── textrecongnition/
│   │── text_detection.py  # Speech-to-text processing using Whisper
│──UI_setup.py  # Tkinter-based voice chat UI
│── .gitignore
│── main.py  # Entry point to run emotion detection
│── requirements.txt  # List of dependencies

Install Dependencies

1. Create & activate a virtual environment (optional but recommended):

   python3 -m venv venv
   source venv/bin/activate  # For Linux/macOS
   venv\Scripts\activate     # For Windows
   

2. Install required dependencies:

   pip install -r requirements.txt
   

How to Use emo_predictor

The function emo_predictor(audio_path) takes an audio file as input and returns a dictionary mapping emotion labels to their respective probabilities.

Example Usage:

from emorecognition.emreco import emo_predictor

audio_file = "emorecognition/m4atestfolder/suprise.m4a"
emotion_probs = emo_predictor(audio_file)

print("Emotion Probabilities:", emotion_probs)

How to Use process_audio for Text Detection

The function process_audio(audio_path) converts speech to text using OpenAI Whisper.

Example Usage:

from textrecongnition.text_detection import process_audio

audio_file = "data/audio_examples/happy_text.m4a"
result = process_audio(audio_file)

print("Transcribed Text:", result["text"])

How to Use the Voice Chat UI

The project includes a Tkinter-based UI that allows users to record voice, analyze emotions, and interact with a chatbot.

Run the UI:

python ui/UI_setup.py

UI Features:

• Start & Stop Recording: Users can record voice and analyze emotions in real time.
• Chat History Display: Recognized text and detected emotions are displayed interactively.
• Exit Button: Easily close the UI.

🚀 Now you can use emo_predictor for emotion detection, process_audio for text detection, and the interactive UI for real-time voice chat analysis! 🎤

For the control group we used the branch 'memory_per_user_no_memory'