aryankargwal/capbot2.0

Deep learning-based image captioning system that processes multiple CCTV feeds in real-time, generating searchable text logs indexed by camera location and timestamp. Built with Streamlit for dual interfaces—one for continuous feed captioning and another for keyword-based incident retrieval with encryption support for stored captions. Reduces manual surveillance overhead by automating footage analysis and enabling rapid incident lookup without reviewing raw video.

No commits in the last 6 months.