Neuron Segmentation Using Incomplete and Noisy Labels Via Adaptive Learning with Structure Priors

Authors
ChanMin ParkKanggeun LeeKim Su YeonFatma Sema Canbakis CecenSeok-Kyu KwonWon-Ki Jeong
Issue Date
2021-04
Publisher
IEEE
Citation
18th IEEE International Symposium on Biomedical Imaging (ISBI), pp.1466 - 1470
Abstract
Recent advances in machine learning have shown significant success in biomedical image segmentation. Most existing high-quality segmentation algorithms rely on supervised learning with full training labels. However, such methods are more susceptible to label quality; besides, generating accurate labels in biomedical data is a labor- and time-intensive task. In this paper, we propose a novel neuron segmentation method that uses only incomplete and noisy labels. The proposed method employs a noise-tolerant adaptive loss that handles partially annotated labels. Moreover, the proposed reconstruction loss leverages prior knowledge of neuronal cell structures to reduce false segmentation near noisy labels. the proposed loss function outperforms several widely used stateof-the-art noise-tolerant losses, such as reverse cross entropy, normalized cross entropy and noise-robust dice losses.
ISSN
1945-7928
URI
https://pubs.kist.re.kr/handle/201004/77761
DOI
10.1109/ISBI48211.2021.9434102
Appears in Collections:
KIST Conference Paper > 2021
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML

qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE