FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3200875269> ?p ?o ?g. }

• W3200875269 abstract "Convolutional Neural Networks (ConvNets) at present achieve remarkable performance in image classification tasks. However, current ConvNets cannot guarantee the capabilities of the mammalian visual systems such as invariance to contrast and illumination changes. Some ideas to overcome the illumination and contrast variations usually have to be tuned manually and tend to fail when tested with other types of data degradation. In this context, we present a new bio-inspired {entry} layer, M6, which detects low-level geometric features (lines, edges, and orientations) which are similar to patterns detected by the V1 visual cortex. This new trainable layer is capable of coping with image classification even with large contrast variations. The explanation for this behavior is the monogenic signal geometry, which represents each pixel value in a 3D space using quaternions, a fact that confers a degree of explainability to the networks. We compare M6 with a conventional convolutional layer (C) and a deterministic quaternion local phase layer (Q9). The experimental setup {is designed to evaluate the robustness} of our M6 enriched ConvNet model and includes three architectures, four datasets, three types of contrast degradation (including non-uniform haze degradations). The numerical results reveal that the models with M6 are the most robust in front of any kind of contrast variations. This amounts to a significant enhancement of the C models, which usually have reasonably good performance only when the same training and test degradation are used, except for the case of maximum degradation. Moreover, the Structural Similarity Index Measure (SSIM) is used to analyze and explain the robustness effect of the M6 feature maps under any kind of contrast degradations." @default.
• W3200875269 created "2021-09-27" @default.
• W3200875269 creator A5039873301 @default.
• W3200875269 creator A5041398333 @default.
• W3200875269 creator A5046887270 @default.
• W3200875269 creator A5050846205 @default.
• W3200875269 creator A5070067140 @default.
• W3200875269 date "2021-09-14" @default.
• W3200875269 modified "2023-10-14" @default.
• W3200875269 title "A trainable monogenic ConvNet layer robust in front of large contrast changes in image classification" @default.
• W3200875269 cites W1242520691 @default.
• W3200875269 cites W1553200558 @default.
• W3200875269 cites W1862131709 @default.
• W3200875269 cites W191734164 @default.
• W3200875269 cites W2003126117 @default.
• W3200875269 cites W2006500012 @default.
• W3200875269 cites W2020898814 @default.
• W3200875269 cites W2034787273 @default.
• W3200875269 cites W2035250029 @default.
• W3200875269 cites W2037552351 @default.
• W3200875269 cites W2046446176 @default.
• W3200875269 cites W2086087329 @default.
• W3200875269 cites W2093559072 @default.
• W3200875269 cites W2105994703 @default.
• W3200875269 cites W2116360511 @default.
• W3200875269 cites W2128254161 @default.
• W3200875269 cites W2135269154 @default.
• W3200875269 cites W2156163116 @default.
• W3200875269 cites W2156749117 @default.
• W3200875269 cites W2163605009 @default.
• W3200875269 cites W2271840356 @default.
• W3200875269 cites W2279221249 @default.
• W3200875269 cites W2557283755 @default.
• W3200875269 cites W2576915720 @default.
• W3200875269 cites W2601564443 @default.
• W3200875269 cites W2737036732 @default.
• W3200875269 cites W2750384547 @default.
• W3200875269 cites W2785885246 @default.
• W3200875269 cites W2807463582 @default.
• W3200875269 cites W2808617845 @default.
• W3200875269 cites W2901406073 @default.
• W3200875269 cites W2914484425 @default.
• W3200875269 cites W2963060032 @default.
• W3200875269 cites W2963676366 @default.
• W3200875269 cites W2963980515 @default.
• W3200875269 cites W2973099873 @default.
• W3200875269 cites W2994264110 @default.
• W3200875269 cites W3008266720 @default.
• W3200875269 cites W3098649723 @default.
• W3200875269 cites W3103272223 @default.
• W3200875269 cites W3110231755 @default.
• W3200875269 cites W3113453806 @default.
• W3200875269 cites W3118608800 @default.
• W3200875269 cites W3183379515 @default.
• W3200875269 cites W584173323 @default.
• W3200875269 cites W2963459876 @default.
• W3200875269 cites W2963543962 @default.
• W3200875269 hasPublicationYear "2021" @default.
• W3200875269 type Work @default.
• W3200875269 sameAs 3200875269 @default.
• W3200875269 citedByCount "0" @default.
• W3200875269 crossrefType "posted-content" @default.
• W3200875269 hasAuthorship W3200875269A5039873301 @default.
• W3200875269 hasAuthorship W3200875269A5041398333 @default.
• W3200875269 hasAuthorship W3200875269A5046887270 @default.
• W3200875269 hasAuthorship W3200875269A5050846205 @default.
• W3200875269 hasAuthorship W3200875269A5070067140 @default.
• W3200875269 hasConcept C104317684 @default.
• W3200875269 hasConcept C153180895 @default.
• W3200875269 hasConcept C154945302 @default.
• W3200875269 hasConcept C160633673 @default.
• W3200875269 hasConcept C178790620 @default.
• W3200875269 hasConcept C185592680 @default.
• W3200875269 hasConcept C200127275 @default.
• W3200875269 hasConcept C2524010 @default.
• W3200875269 hasConcept C2776502983 @default.
• W3200875269 hasConcept C2779227376 @default.
• W3200875269 hasConcept C33923547 @default.
• W3200875269 hasConcept C41008148 @default.
• W3200875269 hasConcept C55493867 @default.
• W3200875269 hasConcept C63479239 @default.
• W3200875269 hasConcept C81363708 @default.
• W3200875269 hasConceptScore W3200875269C104317684 @default.
• W3200875269 hasConceptScore W3200875269C153180895 @default.
• W3200875269 hasConceptScore W3200875269C154945302 @default.
• W3200875269 hasConceptScore W3200875269C160633673 @default.
• W3200875269 hasConceptScore W3200875269C178790620 @default.
• W3200875269 hasConceptScore W3200875269C185592680 @default.
• W3200875269 hasConceptScore W3200875269C200127275 @default.
• W3200875269 hasConceptScore W3200875269C2524010 @default.
• W3200875269 hasConceptScore W3200875269C2776502983 @default.
• W3200875269 hasConceptScore W3200875269C2779227376 @default.
• W3200875269 hasConceptScore W3200875269C33923547 @default.
• W3200875269 hasConceptScore W3200875269C41008148 @default.
• W3200875269 hasConceptScore W3200875269C55493867 @default.
• W3200875269 hasConceptScore W3200875269C63479239 @default.
• W3200875269 hasConceptScore W3200875269C81363708 @default.
• W3200875269 hasLocation W32008752691 @default.
• W3200875269 hasOpenAccess W3200875269 @default.
• W3200875269 hasPrimaryLocation W32008752691 @default.

• next