ThreshNet: Segmentation Refinement Inspired by Region-Specific Thresholding

11/12/2022
by   Savinay Nagendra, et al.
0

We present ThreshNet, a post-processing method to refine the output of neural networks designed for binary segmentation tasks. ThreshNet uses the confidence map produced by a base network along with global and local patch information to significantly improve the performance of even state-of-the-art methods. Binary segmentation models typically convert confidence maps into predictions by thresholding the confidence scores at 0.5 (or some other fixed number). However, we observe that the best threshold is image-dependent and often even region-specific – different parts of the image benefit from using different thresholds. Thus ThreshNet takes a trained segmentation model and learns to correct its predictions by using a memory-efficient post-processing architecture that incorporates region-specific thresholds as part of the training mechanism. Our experiments show that ThreshNet consistently improves over current the state-of-the-art methods in binary segmentation and saliency detection, typically by 3 to 5

READ FULL TEXT

page 1

page 2

page 4

page 7

page 8

research
10/07/2021

Learning post-processing for QRS detection using Recurrent Neural Network

Deep-learning based QRS-detection algorithms often require essential pos...
research
07/21/2018

A post-processing method to improve the white matter hyperintensity segmentation accuracy for randomly-initialized U-net

White matter hyperintensity (WMH) is commonly found in elder individuals...
research
02/21/2018

Semantic Segmentation Refinement by Monte Carlo Region Growing of High Confidence Detections

Despite recent improvements using fully convolutional networks, in gener...
research
02/28/2017

Supervised Saliency Map Driven Segmentation of the Lesions in Dermoscopic Images

Lesion segmentation is the first step in the most automatic melanoma rec...
research
12/28/2022

Building Segmentation on Satellite Images and Performance of Post-Processing Methods

Researchers are doing intensive work on satellite images due to the info...
research
05/12/2020

Probabilistic Semantic Segmentation Refinement by Monte Carlo Region Growing

Semantic segmentation with fine-grained pixel-level accuracy is a fundam...
research
08/07/2017

Training Deep Networks to be Spatially Sensitive

In many computer vision tasks, for example saliency prediction or semant...

Please sign up or login with your details

Forgot password? Click here to reset