Aibu - image sequence annotator

Aibu is an image sequence annotator developed and used by the VOT challenge. User can annotate movement of a single target in an image sequence.

The annotator was only used within the VOT challenge organization until now and is therefore considered an alpha level software. A lot of features are missing or are not fully polished. If you find such a case you are welcome to open an issue report on Github.

Features:

• input: image sequence
• per-frame tags
• bounding box interpolation
• aligned or rotated bounding box
• undo, redo

Building

The annotator is written in Java and can be built using Ant build tool. It depends on the coffeshop library.

$ # build and add coffeshop library to your classpath
$ ant

deploy/aibu.jar will be created

Using

$ java -jar deploy/aibu.jar si.vicos.annotations.editor.Annotator

Sample Annotations

Available on http://votchallenge.net/.