Description
Object detection at full resolution on HD camera feeds is slow, while downscaling blurs the output. CCTV_YOLO shows a middle path: run YOLOv5n6 on low-resolution frames and draw results on the high-resolution frames.
It demonstrates on a public Las Vegas sidewalk traffic cam with a Gradio UI you can try on Hugging Face Spaces.
Fast inference:Low-res detection.
HD output:Boxes on full frames.
Live streams:Real-time video.
Try online:Hugging Face Space.
It demonstrates on a public Las Vegas sidewalk traffic cam with a Gradio UI you can try on Hugging Face Spaces.
Features
Fast inference:Low-res detection.
HD output:Boxes on full frames.
Live streams:Real-time video.
Try online:Hugging Face Space.
