April 2020
tl;dr: Summary of the main idea.
CenterMask works almost in exactly the same way as BlendMask. See my review on Medium.
- CenterMask uses 1 prototype mask (named global saliency map) explicitly.
- CenterMask's name comes from the fact that it uses CenterNet as the backbone, while CenterMask uses the similar anchor-free one-stage FCOS as backbone.
- Summaries of the key ideas
- Summary of technical details
- Questions and notes on how to improve/revise the current work