Detection COCO format is missing

[franva]

From this page, it mentions that for the Detection task, it supports both VOC and COCO, but the format of COCO is not mentioned.

Here is the link: https://github.com/PaddleCV-SIG/PP-Label/wiki/Dataset-Structure#detection

[linhandev]

we are in the process of implementing detection dataste import and export. function and doc will be finallized within two weeks

[franva]

looking forward to seeing it

[linhandev]

Hi @franva, we have recently added support for im/exporting coco format detection datasets. Maybe try it out.