FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4383753493> ?p ?o ?g. }

• W4383753493 endingPage "10" @default.
• W4383753493 startingPage "1" @default.
• W4383753493 abstract "No-service rail surface defect (NRSD) segmentation is an essential way for perceiving the quality of no-service rails. However, due to the complex and diverse outlines and low-contrast textures of no-service rails, existing natural image segmentation methods cannot achieve promising performance in NRSD images, especially in some unique and challenging NRSD scenes, such as low illumination, chaotic background, multiple/tiny defects, and inconsistent defect. To this end, in this paper, we propose a novel segmentation network for NRSDs based on Normalized Attention and Dual-scale Interaction, named NaDiNet. Specifically, NaDiNet follows the enhancement-interaction paradigm. The <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>Normalized Channel-wise Self-Attention Module</i> (NAM) and the <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>Dual-scale Interaction Block</i> (DIB) are two key components of NaDiNet. NAM is a specific extension of the channel-wise self-attention mechanism (CAM) to enhance features extracted from low-contrast NRSD images. The softmax layer in CAM will produce very small correlation coefficients which are not conducive to low-contrast feature enhancement. Instead, in NAM, we directly calculate the normalized correlation coefficient between channels to enlarge the feature differentiation. DIB is specifically designed for the feature interaction of the enhanced features. It has two interaction branches with dual scales, one for fine-grained clues and the other for coarse-grained clues. With both branches working together, DIB can perceive defect regions of different granularities. With these modules working together, our NaDiNet can generate accurate segmentation map. Extensive experiments on the public NRSD-MN dataset with man-made and natural NRSDs demonstrate that our proposed NaDiNet with various backbones ( <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>i.e</i> ., VGG, ResNet, and DenseNet) consistently outperforms 10 state-of-the-art methods. The code and results of our method are available at https://github.com/monxxcn/NaDiNet." @default.
• W4383753493 created "2023-07-11" @default.
• W4383753493 creator A5004105620 @default.
• W4383753493 creator A5040009629 @default.
• W4383753493 creator A5061140967 @default.
• W4383753493 date "2023-01-01" @default.
• W4383753493 modified "2023-09-26" @default.
• W4383753493 title "No-Service Rail Surface Defect Segmentation via Normalized Attention and Dual-scale Interaction" @default.
• W4383753493 cites W1677182931 @default.
• W4383753493 cites W1745334888 @default.
• W4383753493 cites W1903029394 @default.
• W4383753493 cites W2147800946 @default.
• W4383753493 cites W2155903085 @default.
• W4383753493 cites W2194775991 @default.
• W4383753493 cites W2412782625 @default.
• W4383753493 cites W2560023338 @default.
• W4383753493 cites W2799213142 @default.
• W4383753493 cites W2890391196 @default.
• W4383753493 cites W2939355860 @default.
• W4383753493 cites W2955058313 @default.
• W4383753493 cites W2963446712 @default.
• W4383753493 cites W2963685207 @default.
• W4383753493 cites W2963881378 @default.
• W4383753493 cites W2964309882 @default.
• W4383753493 cites W2981689412 @default.
• W4383753493 cites W2998291476 @default.
• W4383753493 cites W3007464067 @default.
• W4383753493 cites W3010616503 @default.
• W4383753493 cites W3035487542 @default.
• W4383753493 cites W3097053213 @default.
• W4383753493 cites W3106587394 @default.
• W4383753493 cites W3113270537 @default.
• W4383753493 cites W3135874576 @default.
• W4383753493 cites W3194222902 @default.
• W4383753493 cites W3204697265 @default.
• W4383753493 cites W3210144980 @default.
• W4383753493 cites W3215156797 @default.
• W4383753493 cites W4206670089 @default.
• W4383753493 cites W4214893857 @default.
• W4383753493 cites W4221138999 @default.
• W4383753493 cites W4281683263 @default.
• W4383753493 cites W4285120968 @default.
• W4383753493 cites W4292347849 @default.
• W4383753493 cites W4296913506 @default.
• W4383753493 cites W4303709937 @default.
• W4383753493 cites W4312226907 @default.
• W4383753493 cites W4313537322 @default.
• W4383753493 cites W4319996167 @default.
• W4383753493 cites W4321380746 @default.
• W4383753493 doi "https://doi.org/10.1109/tim.2023.3293561" @default.
• W4383753493 hasPublicationYear "2023" @default.
• W4383753493 type Work @default.
• W4383753493 citedByCount "0" @default.
• W4383753493 crossrefType "journal-article" @default.
• W4383753493 hasAuthorship W4383753493A5004105620 @default.
• W4383753493 hasAuthorship W4383753493A5040009629 @default.
• W4383753493 hasAuthorship W4383753493A5061140967 @default.
• W4383753493 hasBestOaLocation W43837534932 @default.
• W4383753493 hasConcept C121332964 @default.
• W4383753493 hasConcept C124952713 @default.
• W4383753493 hasConcept C127162648 @default.
• W4383753493 hasConcept C138885662 @default.
• W4383753493 hasConcept C142362112 @default.
• W4383753493 hasConcept C153180895 @default.
• W4383753493 hasConcept C154945302 @default.
• W4383753493 hasConcept C2776401178 @default.
• W4383753493 hasConcept C2776502983 @default.
• W4383753493 hasConcept C2778755073 @default.
• W4383753493 hasConcept C2780980858 @default.
• W4383753493 hasConcept C31258907 @default.
• W4383753493 hasConcept C31972630 @default.
• W4383753493 hasConcept C41008148 @default.
• W4383753493 hasConcept C41895202 @default.
• W4383753493 hasConcept C62520636 @default.
• W4383753493 hasConcept C89600930 @default.
• W4383753493 hasConceptScore W4383753493C121332964 @default.
• W4383753493 hasConceptScore W4383753493C124952713 @default.
• W4383753493 hasConceptScore W4383753493C127162648 @default.
• W4383753493 hasConceptScore W4383753493C138885662 @default.
• W4383753493 hasConceptScore W4383753493C142362112 @default.
• W4383753493 hasConceptScore W4383753493C153180895 @default.
• W4383753493 hasConceptScore W4383753493C154945302 @default.
• W4383753493 hasConceptScore W4383753493C2776401178 @default.
• W4383753493 hasConceptScore W4383753493C2776502983 @default.
• W4383753493 hasConceptScore W4383753493C2778755073 @default.
• W4383753493 hasConceptScore W4383753493C2780980858 @default.
• W4383753493 hasConceptScore W4383753493C31258907 @default.
• W4383753493 hasConceptScore W4383753493C31972630 @default.
• W4383753493 hasConceptScore W4383753493C41008148 @default.
• W4383753493 hasConceptScore W4383753493C41895202 @default.
• W4383753493 hasConceptScore W4383753493C62520636 @default.
• W4383753493 hasConceptScore W4383753493C89600930 @default.
• W4383753493 hasFunder F4320321001 @default.
• W4383753493 hasFunder F4320321543 @default.
• W4383753493 hasLocation W43837534931 @default.
• W4383753493 hasLocation W43837534932 @default.
• W4383753493 hasOpenAccess W4383753493 @default.
• W4383753493 hasPrimaryLocation W43837534931 @default.

• next