DeepAI
Log In Sign Up

Learning Non-Unique Segmentation with Reward-Penalty Dice Loss

09/23/2020
by   Jiabo He, et al.
0

Semantic segmentation is one of the key problems in the field of computer vision, as it enables computer image understanding. However, most research and applications of semantic segmentation focus on addressing unique segmentation problems, where there is only one gold standard segmentation result for every input image. This may not be true in some problems, e.g., medical applications. We may have non-unique segmentation annotations as different surgeons may perform successful surgeries for the same patient in slightly different ways. To comprehensively learn non-unique segmentation tasks, we propose the reward-penalty Dice loss (RPDL) function as the optimization objective for deep convolutional neural networks (DCNN). RPDL is capable of helping DCNN learn non-unique segmentation by enhancing common regions and penalizing outside ones. Experimental results show that RPDL improves the performance of DCNN models by up to 18.4 surgical dataset.

READ FULL TEXT

page 1

page 4

page 6

09/11/2018

Normalization in Training Deep Convolutional Neural Networks for 2D Bio-medical Semantic Segmentation

2D bio-medical semantic segmentation is important for surgical robotic v...
08/10/2019

Distance Map Loss Penalty Term for Semantic Segmentation

Convolutional neural networks for semantic segmentation suffer from low ...
05/25/2021

Review on Indoor RGB-D Semantic Segmentation with Deep Convolutional Neural Networks

Many research works focus on leveraging the complementary geometric info...
01/13/2020

Evolution of Image Segmentation using Deep Convolutional Neural Network: A Survey

From the autonomous car driving to medical diagnosis, the requirement of...
02/26/2017

Adversarial Networks for the Detection of Aggressive Prostate Cancer

Semantic segmentation constitutes an integral part of medical image anal...
02/23/2020

Random Bundle: Brain Metastases Segmentation Ensembling through Annotation Randomization

We introduce a novel ensembling method, Random Bundle (RB), that improve...