Video Task Types

Note: Scale Video is only available for our Enterprise customers. If you want to learn more, please contact our sales team.

Scale offers two distinct video task types: videoannotation and videoplaybackannotation. Both video task types support annotation of multiple geometries (box, polygon, line, point, cuboid, and ellipse) in a single task. For performance reasons, we recommend using videoplaybackannotation if the video you want labeled has more than 100 frames.

Here is a detailed comparison:

videoannotation

videoplaybackannotation

Supported attachments

An array of image frames or a single video file

An array of image frames or a single video file

Labeling interface

Video displayed as sequence of image frames

Video displayed as embedded video

Video length

Works well with videos of <100 frames

Works well with videos of up to tens of thousands of frames