devprincekumar/yawcam-ai repository overview

A ready-to-run edge AI container that delivers YOLO-powered detection, recording, and event automation with a headless web UI, RTSP support, persistent storage, and optional CUDA acceleration.

The CPU image runs anywhere (laptops, servers, Pi streams), while the GPU build auto-leverages NVIDIA cards for faster inference and falls back to CPU if absent.

Designed for home/industrial surveillance, smart camera pipelines, RTSP NVR workloads, and edge deployments, it supports model swapping for retrained YOLO variants and integrates cleanly with PiStream-Lite for video input and EdgePulse for system health and thermal optimization. Perfect for users who want a portable AI video engine without manual setup, detect, record, automate, and extend directly at the edge.