Panoptic segmentation with highly imbalanced semantic labels

2022/03/03 by Rumberger, Josef Lorenz, Baumann, Elias, Hirsch, Peter +3 · 2 citations
#Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #Image and Video Processing (eess.IV) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.2203.11692

Abstract

We describe here the panoptic segmentation method we devised for our participation in the CoNIC: Colon Nuclei Identification and Counting Challenge at ISBI 2022. Key features of our method are a weighted loss specifically engineered for semantic segmentation of highly imbalanced cell types, and a state-of-the art nuclei instance segmentation model, which we combine in a Hovernet-like architecture.

Cited by

• KongNet: A Multi-headed Deep Learning Model for Detection and Classification of Nuclei in Histopathology Images
• PhenoBench: A Comprehensive Benchmark for Cell Phenotyping

Related

• Nuclei panoptic segmentation and composition regression with multi-task deep neural networks
• Simultaneous Semantic and Instance Segmentation for Colon Nuclei Identification and Counting
• CoNIC: Colon Nuclei Identification and Counting Challenge 2022
• Ensembling Instance and Semantic Segmentation for Panoptic Segmentation
• ciscNet -- A Single-Branch Cell Instance Segmentation and Classification Network