LumiScan Ecosystem: LumiScan Engineering
LumiScan Annotation Web
LumiScan Annotation Web creates the training data for neural networks that detect and locate objects, for example for LumiScan AI Object Matching or LumiScan VGR. It is the browser version and successor of the desktop tool LumiScan Annotation: a server on Windows or Linux holds the images and annotations, and annotators work in their web browser from any PC in the company network, without installing anything.
Several annotators can annotate the same image independently. The server compares their boxes, writes the reference annotation when they agree and shows reviewers where they differ. Box suggestions from your trained networks speed up the work, and the annotation files are exactly those of the desktop tool, so LumiScan Training reads them unchanged.
Annotating
- Oriented boxes as center-line rectangles in three clicks – moved, resized and rotated with mouse or keyboard
- Magnifier with crosshair for edges to a tenth of a pixel
- Labels on keys 1 to 9, brightness and contrast, region of interest
- For image series: boxes from the previous image, copy and paste, box templates per label
- Undo and redo, and a history of every saved state of every image
Working as a team
- Projects with their own images, settings and members
- Roles: annotator, reviewer, project admin and server admin
- Sign-in with local accounts, Active Directory / LDAP or single sign-on (OpenID Connect)
- Edit locks and revision notes from reviewers to annotators
Quality and speed
- Several annotators per image, each without seeing the others' boxes
- Automatic reference annotation when all agree; numbered differences for reviewers
- Agreement statistics per annotator, per pair and per folder
- Box suggestions from your trained LumiScan networks (ONNX) or an open-vocabulary detector
- Active learning: the images the network is least sure about come first
Data and operation
- CSV and JSON files byte-identical to the desktop tool – read unchanged by LumiScan Training
- Projects in folders on the server, on a network share or in an S3 bucket
- PNG, JPEG, BMP and TIFF images, including multi-page TIFF
- Server on Windows 11 or Ubuntu 24.04; users need only Chrome, Edge or Firefox
- Runs entirely in your network – images never leave the company