VITA_reproduction

A code reproduction of VITA.

Installation

1. see Mask2Former
2. see IFC
3. The Datasets need to be placed up the root of detertrons as follows ：catalogue_desc
4. You need to add YouTubeVIS 2019 model weight of Mask2Former about Video Instance Segmentation to weights folder as for as a pre-training model.

Note

• Based on detectron2.
• The codes are under projects/ folder, which follows the convention of detectron2.
• You can easily import our project to the latest detectron2 by following below.
• inserting projects/VITA folder