FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4288049153> ?p ?o ?g. }

• W4288049153 endingPage "1933" @default.
• W4288049153 startingPage "1919" @default.
• W4288049153 abstract "In this paper, we propose a GraphBit method to learn unsupervised deep binary descriptors for efficient image representation. Conventional binary representation learning methods directly quantize each element according to the threshold without considering the quantization ambiguousness. The elements near the boundary dubbed as ambiguous bits fail to collect effective information for reliable binarization and are sensitive to noise that causes reversed bits. We argue that there are implicit inner relationships among bits in binary descriptors called bitwise interaction, where the related bits can provide extra instruction as prior knowledge for ambiguousness reduction. Specifically, we design a deep reinforcement learning model to learn the structure of the graph for bitwise interaction mining, and the uncertainty of binary codes is reduced by maximizing the mutual information with input and related bits. Consequently, the ambiguous bits receive additional instruction from the graph for reliable binarization. Moreover, we further present a differentiable search method (GraphBit+) that mines the bitwise interaction in continuous space, so that the heavy search cost caused by the training difficulties in reinforcement learning is significantly reduced. Since the GraphBit and GraphBit+ methods learn fixed bitwise interaction which is suboptimal for various input, the inaccurate instruction from the fixed bitwise interaction cannot effectively decrease the ambiguousness of binary descriptors. To address this, we further propose the unsupervised binary descriptor learning method via dynamic bitwise interaction mining (D-GraphBit), where a graph convolutional network called GraphMiner reasons the optimal bitwise interaction for each input sample. Extensive experimental results on the CIFAR-10, NUS-WIDE, ImageNet-100, Brown and HPatches datasets demonstrate the efficiency and effectiveness of the proposed GraphBit, GraphBit+ and D-GraphBit." @default.
• W4288049153 created "2022-07-27" @default.
• W4288049153 creator A5010357390 @default.
• W4288049153 creator A5013973037 @default.
• W4288049153 creator A5045134787 @default.
• W4288049153 creator A5055401170 @default.
• W4288049153 creator A5090079801 @default.
• W4288049153 date "2023-02-01" @default.
• W4288049153 modified "2023-10-12" @default.
• W4288049153 title "Learning Deep Binary Descriptors via Bitwise Interaction Mining" @default.
• W4288049153 cites W1491719799 @default.
• W4288049153 cites W1689727326 @default.
• W4288049153 cites W1956333070 @default.
• W4288049153 cites W1974647172 @default.
• W4288049153 cites W1979931042 @default.
• W4288049153 cites W1995266040 @default.
• W4288049153 cites W2007972815 @default.
• W4288049153 cites W2047186200 @default.
• W4288049153 cites W2089888558 @default.
• W4288049153 cites W2097934584 @default.
• W4288049153 cites W2104978738 @default.
• W4288049153 cites W2108598243 @default.
• W4288049153 cites W2117228865 @default.
• W4288049153 cites W2119717200 @default.
• W4288049153 cites W2126338221 @default.
• W4288049153 cites W2134514757 @default.
• W4288049153 cites W2141584146 @default.
• W4288049153 cites W2145065594 @default.
• W4288049153 cites W2145339207 @default.
• W4288049153 cites W2151103935 @default.
• W4288049153 cites W2321533354 @default.
• W4288049153 cites W2325939864 @default.
• W4288049153 cites W2461086877 @default.
• W4288049153 cites W2464915613 @default.
• W4288049153 cites W2493727926 @default.
• W4288049153 cites W2737652487 @default.
• W4288049153 cites W2781821509 @default.
• W4288049153 cites W2798840456 @default.
• W4288049153 cites W2799171885 @default.
• W4288049153 cites W2799245683 @default.
• W4288049153 cites W2913932916 @default.
• W4288049153 cites W2916781626 @default.
• W4288049153 cites W2948246283 @default.
• W4288049153 cites W2963319519 @default.
• W4288049153 cites W2963748588 @default.
• W4288049153 cites W2963791342 @default.
• W4288049153 cites W2964280870 @default.
• W4288049153 cites W2981199548 @default.
• W4288049153 cites W2982892652 @default.
• W4288049153 cites W2987672160 @default.
• W4288049153 cites W3008062779 @default.
• W4288049153 cites W3034275286 @default.
• W4288049153 cites W3034887213 @default.
• W4288049153 cites W3035524453 @default.
• W4288049153 cites W3035559192 @default.
• W4288049153 cites W3036967712 @default.
• W4288049153 cites W3089135124 @default.
• W4288049153 cites W3108075360 @default.
• W4288049153 cites W3188212618 @default.
• W4288049153 cites W343636949 @default.
• W4288049153 doi "https://doi.org/10.1109/tpami.2022.3161600" @default.
• W4288049153 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/35320086" @default.
• W4288049153 hasPublicationYear "2023" @default.
• W4288049153 type Work @default.
• W4288049153 citedByCount "2" @default.
• W4288049153 countsByYear W42880491532023 @default.
• W4288049153 crossrefType "journal-article" @default.
• W4288049153 hasAuthorship W4288049153A5010357390 @default.
• W4288049153 hasAuthorship W4288049153A5013973037 @default.
• W4288049153 hasAuthorship W4288049153A5045134787 @default.
• W4288049153 hasAuthorship W4288049153A5055401170 @default.
• W4288049153 hasAuthorship W4288049153A5090079801 @default.
• W4288049153 hasConcept C134765980 @default.
• W4288049153 hasConcept C153180895 @default.
• W4288049153 hasConcept C154945302 @default.
• W4288049153 hasConcept C199360897 @default.
• W4288049153 hasConcept C33923547 @default.
• W4288049153 hasConcept C41008148 @default.
• W4288049153 hasConcept C48372109 @default.
• W4288049153 hasConcept C63435697 @default.
• W4288049153 hasConcept C80444323 @default.
• W4288049153 hasConcept C94375191 @default.
• W4288049153 hasConceptScore W4288049153C134765980 @default.
• W4288049153 hasConceptScore W4288049153C153180895 @default.
• W4288049153 hasConceptScore W4288049153C154945302 @default.
• W4288049153 hasConceptScore W4288049153C199360897 @default.
• W4288049153 hasConceptScore W4288049153C33923547 @default.
• W4288049153 hasConceptScore W4288049153C41008148 @default.
• W4288049153 hasConceptScore W4288049153C48372109 @default.
• W4288049153 hasConceptScore W4288049153C63435697 @default.
• W4288049153 hasConceptScore W4288049153C80444323 @default.
• W4288049153 hasConceptScore W4288049153C94375191 @default.
• W4288049153 hasFunder F4320321001 @default.
• W4288049153 hasFunder F4320322392 @default.
• W4288049153 hasIssue "2" @default.
• W4288049153 hasLocation W42880491531 @default.
• W4288049153 hasLocation W42880491532 @default.
• W4288049153 hasOpenAccess W4288049153 @default.

• next