I would like to train an instance segmentation model on a custom dataset, for which I converted the annotations in COCO format. Which frameworks would you suggest me to use? I know about torchvision, MMDetection and Detectron2. I need a framework which
- supports instance segmentation (clearly )
- consumes COCO annotations
- is a good compromise between flexibility and robustness/ease of use.
- has a Docker image. Bonus points if the Docker image is available on https://ngc.nvidia.com/, or if itβs compatible with the NVIDIA PyTorch image available there.
Which one would you suggest to use? Thanks,
Andrea