VisionAI is a cross-platform, Python-based smart camera application that delivers real-time AI-powered features in a responsive GUI. Built with YOLOv8, OpenCV, and MediaPipe, it offers seamless object detection, face recognition with privacy blurring, motion tracking, gesture support, and voice commands.
Experience live AI detection: GitHub Repo
- 🔍 Ultra-Fast Object Detection – YOLOv8 Nano inference at 640×640 resolution
- 👁️ Face Recognition & Privacy Blurring – Haar Cascade detection with optional blur for anonymity
- 🏃 Motion Tracking & Alerts – Background subtraction highlights moving objects in real time
- 🤚 Gesture & Pose Estimation (Pro Mode) – Hand and body tracking via MediaPipe
- 🎤 Voice Commands – Control start/stop, recording, and detection toggles with speech
- 🖼️ Recording & Screenshots – Timestamped AVI videos and JPEG snapshots
- 📊 Event Logging & CSV Export – Automated detection logs for analysis
- ⚡ Dual GUI Modes – Optimized Mode for lightweight use, Pro Mode for advanced analytics
- 📦 Easy Distribution – Create standalone EXE or portable ZIP package
📹 Watch Full Demo – Explore VisionAI features: live object detection, face blurring, motion alerts, and the Pro analytics dashboard.
Demo Highlights:
- 🔍 Real-time detection of multiple object classes
- 👁️ Privacy blurring of faces on demand
- 🏃 Highlighted motion regions with area-based alerts
- 🤚 Gesture controls and pose skeleton overlay
- 📊 Exportable CSV log from the analytics panel
# Clone repository
git clone https://github.com/mahimapaseda/VisionAI.git
cd VisionAI
# Install dependencies
pip install -r requirements.txt
# Setup (downloads YOLO model & creates folders)
python setup.py
# Launch application
python vision_ai.pyDistributed under the MIT License. See LICENSE for details.