Preset method

HTC

Introduction

HTC is a multi-task and multi-stage hybrid cascade model designed for object detection and instance segmentation tasks, based on Mask R-CNN and Cascade R-CNN. It combines the characteristics of object detection and instance segmentation. Through the cascade and optimization of multiple stages, it can accurately detect target objects and perform instance segmentation on them simultaneously. This method demonstrates good performance when dealing with instance segmentation tasks in complex scenarios.

Citation

@inproceedings{chen2019hybrid,
  title={Hybrid task cascade for instance segmentation},
  author={Chen, Kai and Pang, Jiangmiao and Wang, Jiaqi and Xiong, Yu and Li, Xiaoxiao and Sun, Shuyang and Feng, Wansen and Liu, Ziwei and Shi, Jianping and Ouyang, Wanli and others},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={4974--4983},
  year={2019}
}

Mask2former

Introduction

Mask2Former is a Transformer-based general segmentation framework that achieves high-performance instance segmentation goals through a concise architecture design and a self-attention mechanism.

Citation

@inproceedings{cheng2022masked,
  title={Masked-attention mask transformer for universal image segmentation},
  author={Cheng, Bowen and Misra, Ishan and Schwing, Alexander G and Kirillov, Alexander and Girdhar, Rohit},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={1290--1299},
  year={2022}
}