Vision Studio – YOLO11, Depth Anything & SegFormer in your browser

Task

YOLO11 object detection — 80 COCO classes with bounding boxes.

Input

…or drag & drop / paste anywhere. Videos play through the model frame by frame.

Model

Confidence 0.25 IoU (NMS) 0.45 Show labels Track objects (IDs + trails, video/webcam) Count line crossings (draw a line on the video) Estimate distance per box (runs Depth Anything too) Fine-grained labels per box (runs classifier too)

Overlay opacity 0.5

Custom model

Any Ultralytics ONNX export (detect / segment / pose / obb / classify). Class names and input size are read from the file.

Model zoo

Export

Advanced

Backend HF model precision Camera resolution

Idle

Drop an image or video here or pick a sample on the left

– ms inference – ms total – fps

Results