An interactive web application that simulates a mock interview using AI. The application features a voice-enabled AI interviewer, provides real-time feedback on verbal answers, and analyzes the user's on-camera body language for emotional cues.
- Interactive UI: A full-screen, split-panel interface with a talking AI avatar and a live user video feed.
- Voice-Enabled AI: The interviewer asks questions out loud using browser-based Text-to-Speech (TTS).
- Dynamic Sessions: Choose between "Technical" and "HR" domains, with 5 random questions asked per session.
- Answer Scoring: Uses a BERT-based model to score answers based on semantic similarity to expected keywords.
- Body Language Analysis: Utilizes OpenCV and DeepFace to analyze the user's video for dominant emotions and provides feedback on confidence and composure.
- Robust Input Handling: Includes checks for empty answers and non-English responses, prompting the user to try again.
- Backend: Python, Flask
- Frontend: HTML, CSS, JavaScript
- Speech-to-Text: OpenAI's Whisper
- NLP Model: PyTorch, Hugging Face Transformers (BERT)
- Computer Vision: OpenCV, DeepFace
-
Clone the repository:
git https://github.com/Mugilan1309/Mock_Interview_AI.git cd Mock_Interview_AI -
Install Python dependencies:
pip install -r requirements.txt
-
Install FFmpeg:
- This project requires FFmpeg. Ensure it is installed on your system and accessible in your system's PATH. You can download it from ffmpeg.org.
-
Run the application:
python app.py
-
Open your web browser and navigate to
http://127.0.0.1:5000.