First Alpha Release

Auto Disclaimer Adder v1.0.0-alpha

Release Highlights

• First official release of our AI-powered smoking scene detection and disclaimer addition tool
• Release Date: 03/03/2025

Key Features

• Automated smoking detection using Moondream2 vision model
• Automatic disclaimer overlay on detected frames
• User-friendly Gradio web interface
• Multiple detection methods (point, detect, query)
• Custom disclaimer image support
• GPU acceleration for faster processing

Technical Details

• Processing speed: ~4 min for 4-min video at 5 FPS
• Supports common video formats (MP4, AVI, MOV)
• Requires Python 3.8+, 8GB RAM (GPU recommended)

Installation

• Docker container available
• Manual installation option with detailed instructions

Known Limitations

• Resource-intensive for very long videos
• Some smoking gestures may not be detected with 100% accuracy

Next Steps

• Improved detection accuracy
• Drinking scene detection
• Batch processing capabilities