@brian-h-wang hi thanks for sharing this wonderful code , i have few queries
can we perform the inference on CPU , is there CPU code available
2.any instance segmenation or sematic segmentation model output can be used ?
3.what is the performance and accuracy of the of the work
The current version of the algorithm is GPU-only, but the "version1" branch has an older version which has slightly worse performance but can run on CPU.
Yes, Mask-RCNN can be switched out for any instance/semantic segmentation model.
Please see the paper linked from the readme for this info
@brian-h-wang hi thanks for sharing this wonderful code , i have few queries
Thanks in advance