vocalpy / crowsetta

A tool to work with any format for annotating vocalizations
https://crowsetta.readthedocs.io/en/latest/
BSD 3-Clause "New" or "Revised" License
49 stars 3 forks source link

ENH: add `'generic-bbox'` format #192

Open NickleDave opened 2 years ago

NickleDave commented 2 years ago

analogous to 'generic-seq'

to represent multiple annotations per file I guess we will need to use something like .json? Or could it just be a .csv?

My first thought was .json because that's what things like COCO and Pascal VOC use for object detection datasets but .json is probably easier