2020
DOI: 10.1098/rspa.2019.0841
|View full text |Cite
|
Sign up to set email alerts
|

Shearlets as feature extractor for semantic edge detection: the model-based and data-driven realm

Abstract: Semantic edge detection has recently gained a lot of attention as an image-processing task, mainly because of its wide range of real-world applications. This is based on the fact that edges in images contain most of the semantic information. Semantic edge detection involves two tasks, namely pure edge detection and edge classification. Those are in fact fundamentally distinct in terms of the level of abstraction that each task requires. This fact is known as the distracted supervision paradox and limits the po… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
4
0

Year Published

2021
2021
2023
2023

Publication Types

Select...
4
2
1

Relationship

0
7

Authors

Journals

citations
Cited by 7 publications
(4 citation statements)
references
References 36 publications
(125 reference statements)
0
4
0
Order By: Relevance
“…The main benefit of adding these skip connections in the network is that if any of the layers impair the performance of the architecture, it is skipped through regularisation. ResNet [16] is a type of network‐in‐network (NIN) architecture as it consists many residual units stacked together. The collection of these residual units forms a building block for the ResNet architecture.…”
Section: Proposed Methodologymentioning
confidence: 99%
See 1 more Smart Citation
“…The main benefit of adding these skip connections in the network is that if any of the layers impair the performance of the architecture, it is skipped through regularisation. ResNet [16] is a type of network‐in‐network (NIN) architecture as it consists many residual units stacked together. The collection of these residual units forms a building block for the ResNet architecture.…”
Section: Proposed Methodologymentioning
confidence: 99%
“…The main benefit of adding these skip connections in the network is that if any of the layers impair the performance of the architecture, it is skipped through regularisation. ResNet [16] F I G U R E 2 InceptionV1 model architecture [16].…”
Section: Inceptionv3mentioning
confidence: 99%
“…Quantitative analysis. As a quantitative measure of the accuracy of our results we use the F 1 score on the image gradient, as is done in [40,2] (see also the DICE metric [30]). The gradient (or edge) F 1 score is a measure of how well we have recovered the gradient image ∇f (i.e.…”
Section: 2mentioning
confidence: 99%
“…A shearlet system is composed of three operations: Dilation, Shearing, and Translation. Here, we briefly review the definition provided by Andrade-Loarca et al (2020).…”
Section: Shearlet Systemsmentioning
confidence: 99%