Skip to content

Latest commit

 

History

History
18 lines (16 loc) · 786 Bytes

README.md

File metadata and controls

18 lines (16 loc) · 786 Bytes
Raw

Mask RCNN in PyTorch

This is a PyTorch implementation of Mask RCNN. This project attempts to

  1. combine Fast Mask RCNN and faster_rcnn_pytorch (Special thanks awesome these authors)
  2. provide experimental setup for researchers to test their new models on top of this framework
  3. build ROIAlign layer in pure PyTorch.

See you in two weeks. (Hopefully, one week after Fast Mask RCNN is done)

Progress

  • Adapt ROIAlign provided in FastMaskRNN
  • ROIAlign layer in pure PyTorch
  • Training
  • Evaluation
  • COCO Data Provider
  • Finishing writing this list
  • Checking the checkboxes