Drag & Drop Upload
Drop any JPEG, PNG, or WebP image straight onto the canvas workspace.
Create unlimited class labels, draw multiple bounding boxes per image, and export multi-object Pascal VOC XML — ready to feed your Python training pipeline with zero installs.
Drop any JPEG, PNG, or WebP image straight onto the canvas workspace.
Click and drag to draw pixel-accurate rectangular annotations with live coordinate feedback.
One-click download of fully compliant Pascal VOC XML with one <object> block per annotation — plug straight into your training scripts.
Add, rename, delete, and colour-code unlimited class labels on the fly. Each label gets its own colour — visible on the canvas and in the exported XML.
Ships with a mock POST /annotate endpoint — swap in your real FastAPI backend in seconds.
Your images never leave the browser. All annotation logic runs locally for maximum privacy.
Define your labels, draw boxes for each class, export a multi-object Pascal VOC XML.
Drop image here
or click to browse — JPEG, PNG, WebP