Sentinel AI is a video anomaly detection and narration application Users can upload videos for training or detection:
- During training, the system learns normal video behavior using Autoencoder and LSTM models.
- During detection, it identifies any anomalies in the uploaded video and automatically generates a narration describing the detected anomalies.
This enables smart monitoring, helping users understand unusual patterns or activities in videos through AI-driven narration.
✅ Upload videos for training or anomaly detection
✅ Uses Autoencoder and LSTM models for anomaly identification
✅ Automatically generates narrations for detected anomalies
✅ Interactive web interface built with ReactJS
✅ Backend powered by Flask (Python)
✅ Containerized using Docker Compose for seamless deployment
✅ Scalable architecture supporting model retraining and multi-video processing
