SOLO: Segmenting Objects by Locations
Xinlong Wang1Tao Kong2Chunhua Shen1Yuning Jiang2Lei Li2
1The University of Adelaide,    2ByteDance AI Lab
SOLO is totally box-free instance segmentation framework thus not being restricted by (anchor) box locations and scales, and naturally benefits from the inherent advantages of FCNs. Our method takes an image as input, directly outputs instance masks and corresponding class probabilities, in a fully convolutional, box-free and grouping-free paradigm.
Visualization of instance segmentation results using the Res-101-FPN backbone. The model is trained on the COCO train2017 dataset, achieving a mask AP of 37.8 on the COCO test-dev.
SOLO is also able to perform r instance contour detection.
  title     =  {{SOLO}: Segmenting Objects by Locations},
  author    =  {Wang, Xinlong and Kong, Tao and Shen, Chunhua and Jiang, Yuning and Li, Lei},
  booktitle =  {Proc. Eur. Conf. Computer Vision (ECCV)},
  year      =  {2020}
  title={SOLOv2: Dynamic and Fast Instance Segmentation},
  author={Wang, Xinlong and Zhang, Rufeng and  Kong, Tao and Li, Lei and Shen, Chunhua},
  journal={Proc. Advances in Neural Information Processing Systems (NeurIPS)},