Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313153605> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W4313153605 abstract "Image hashing is a principled approximate nearest neighbor approach to find similar items to a query in a large collection of images. Hashing aims to learn a binary-output function that maps an image to a binary vector. For optimal retrieval performance, producing balanced hash codes with low-quantization error to bridge the gap between the learning stage's continuous relaxation and the inference stage's discrete quantization is important. However, in the existing deep supervised hashing methods, coding balance and low-quantization error are difficult to achieve and involve several losses. We argue that this is because the existing quantization approaches in these methods are heuristically constructed and not effective to achieve these objectives. This paper considers an alternative approach to learning the quantization constraints. The task of learning balanced codes with low quantization error is re-formulated as matching the learned distribution of the continuous codes to a pre-defined discrete, uniform distribution. This is equivalent to minimizing the distance between two distributions. We then propose a computationally efficient distributional distance by leveraging the discrete property of the hash functions. This distributional distance is a valid distance and enjoys lower time and sample complexities. The proposed single-loss quantization objective can be integrated into any existing supervised hashing method to improve code balance and quantization error. Experiments confirm that the proposed approach substantially improves the performance of several representative hashing methods." @default.
- W4313153605 created "2023-01-06" @default.
- W4313153605 creator A5067923548 @default.
- W4313153605 creator A5073070877 @default.
- W4313153605 creator A5080642445 @default.
- W4313153605 date "2022-06-01" @default.
- W4313153605 modified "2023-10-16" @default.
- W4313153605 title "One Loss for Quantization: Deep Hashing with Discrete Wasserstein Distributional Matching" @default.
- W4313153605 cites W1910300841 @default.
- W4313153605 cites W1913628733 @default.
- W4313153605 cites W1974647172 @default.
- W4313153605 cites W1982193706 @default.
- W4313153605 cites W1985123706 @default.
- W4313153605 cites W1992371516 @default.
- W4313153605 cites W2029205712 @default.
- W4313153605 cites W2055906546 @default.
- W4313153605 cites W2140431670 @default.
- W4313153605 cites W2152565070 @default.
- W4313153605 cites W2293824885 @default.
- W4313153605 cites W2411707397 @default.
- W4313153605 cites W2461086877 @default.
- W4313153605 cites W2509619282 @default.
- W4313153605 cites W2519998487 @default.
- W4313153605 cites W2586937979 @default.
- W4313153605 cites W2594589367 @default.
- W4313153605 cites W2798834175 @default.
- W4313153605 cites W2798956329 @default.
- W4313153605 cites W2799245683 @default.
- W4313153605 cites W2808282156 @default.
- W4313153605 cites W2913932916 @default.
- W4313153605 cites W2963398989 @default.
- W4313153605 cites W2963517218 @default.
- W4313153605 cites W2964158883 @default.
- W4313153605 cites W2964280870 @default.
- W4313153605 cites W2979557588 @default.
- W4313153605 cites W2986340518 @default.
- W4313153605 cites W3012800455 @default.
- W4313153605 cites W3020188837 @default.
- W4313153605 cites W3034239448 @default.
- W4313153605 cites W4230940751 @default.
- W4313153605 cites W89155778 @default.
- W4313153605 doi "https://doi.org/10.1109/cvpr52688.2022.00923" @default.
- W4313153605 hasPublicationYear "2022" @default.
- W4313153605 type Work @default.
- W4313153605 citedByCount "8" @default.
- W4313153605 countsByYear W43131536052023 @default.
- W4313153605 crossrefType "proceedings-article" @default.
- W4313153605 hasAuthorship W4313153605A5067923548 @default.
- W4313153605 hasAuthorship W4313153605A5073070877 @default.
- W4313153605 hasAuthorship W4313153605A5080642445 @default.
- W4313153605 hasBestOaLocation W43131536052 @default.
- W4313153605 hasConcept C11413529 @default.
- W4313153605 hasConcept C153180895 @default.
- W4313153605 hasConcept C154945302 @default.
- W4313153605 hasConcept C199833920 @default.
- W4313153605 hasConcept C2776214188 @default.
- W4313153605 hasConcept C28855332 @default.
- W4313153605 hasConcept C33923547 @default.
- W4313153605 hasConcept C38652104 @default.
- W4313153605 hasConcept C41008148 @default.
- W4313153605 hasConcept C48372109 @default.
- W4313153605 hasConcept C63435697 @default.
- W4313153605 hasConcept C80444323 @default.
- W4313153605 hasConcept C94375191 @default.
- W4313153605 hasConcept C99138194 @default.
- W4313153605 hasConceptScore W4313153605C11413529 @default.
- W4313153605 hasConceptScore W4313153605C153180895 @default.
- W4313153605 hasConceptScore W4313153605C154945302 @default.
- W4313153605 hasConceptScore W4313153605C199833920 @default.
- W4313153605 hasConceptScore W4313153605C2776214188 @default.
- W4313153605 hasConceptScore W4313153605C28855332 @default.
- W4313153605 hasConceptScore W4313153605C33923547 @default.
- W4313153605 hasConceptScore W4313153605C38652104 @default.
- W4313153605 hasConceptScore W4313153605C41008148 @default.
- W4313153605 hasConceptScore W4313153605C48372109 @default.
- W4313153605 hasConceptScore W4313153605C63435697 @default.
- W4313153605 hasConceptScore W4313153605C80444323 @default.
- W4313153605 hasConceptScore W4313153605C94375191 @default.
- W4313153605 hasConceptScore W4313153605C99138194 @default.
- W4313153605 hasLocation W43131536051 @default.
- W4313153605 hasLocation W43131536052 @default.
- W4313153605 hasOpenAccess W4313153605 @default.
- W4313153605 hasPrimaryLocation W43131536051 @default.
- W4313153605 hasRelatedWork W1956333070 @default.
- W4313153605 hasRelatedWork W2303614492 @default.
- W4313153605 hasRelatedWork W2608884419 @default.
- W4313153605 hasRelatedWork W2737652487 @default.
- W4313153605 hasRelatedWork W2884006160 @default.
- W4313153605 hasRelatedWork W2950261206 @default.
- W4313153605 hasRelatedWork W3199768404 @default.
- W4313153605 hasRelatedWork W4226019763 @default.
- W4313153605 hasRelatedWork W4285052569 @default.
- W4313153605 hasRelatedWork W4376224613 @default.
- W4313153605 isParatext "false" @default.
- W4313153605 isRetracted "false" @default.
- W4313153605 workType "article" @default.