Matches in SemOpenAlex for { <https://semopenalex.org/work/W3129072783> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W3129072783 endingPage "108528" @default.
- W3129072783 startingPage "108528" @default.
- W3129072783 abstract "Zero-shot sketch-based image retrieval (ZS-SBIR) has recently attracted the attention of the computer vision community due to its real-world applications, and the more realistic and challenging setting that it presents over SBIR. ZS-SBIR inherits the main challenges of multiple computer vision problems including content-based Image Retrieval (CBIR), zero-shot learning and domain adaptation. The majority of previous studies using deep neural networks have achieved improved results by either projecting sketch and images into a common low-dimensional space, or transferring knowledge from seen to unseen classes. However, those approaches are trained with complex frameworks composed of multiple deep convolutional neural networks (CNNs) and are dependent on category-level word labels. This increases the requirements for training resources and datasets. In comparison, we propose a simple and efficient framework that does not require high computational training resources, and learns the semantic embedding space from a vision model rather than a language model, as is done by related studies. Furthermore, at training and inference stages our method only uses a single CNN. In this work, a pre-trained ImageNet CNN (i.e., ResNet50) is fine-tuned with three proposed learning objects: domain-balanced quadruplet loss, semantic classification loss, and semantic knowledge preservation loss. The domain-balanced quadruplet and semantic classification losses are introduced to learn discriminative, semantic and domain invariant features by considering ZS-SBIR as an object detection and verification problem. To preserve semantic knowledge learned with ImageNet and exploit it for unseen categories, the semantic knowledge preservation loss is proposed. To reduce computational cost and increase the accuracy of the semantic knowledge distillation process, ground-truth semantic knowledge is prepared in a class-oriented fashion prior to training. Extensive experiments are conducted on three challenging ZS-SBIR datasets: Sketchy Extended, TU-Berlin Extended and QuickDraw Extended. The proposed method achieves state-of-the-art results, and outperforms the majority of related works by a substantial margin." @default.
- W3129072783 created "2021-02-15" @default.
- W3129072783 creator A5034095159 @default.
- W3129072783 creator A5055128383 @default.
- W3129072783 creator A5057063594 @default.
- W3129072783 creator A5080301434 @default.
- W3129072783 creator A5083626840 @default.
- W3129072783 date "2022-06-01" @default.
- W3129072783 modified "2023-10-16" @default.
- W3129072783 title "An efficient framework for zero-shot sketch-based image retrieval" @default.
- W3129072783 cites W1974647172 @default.
- W3129072783 cites W2128543433 @default.
- W3129072783 cites W2466618734 @default.
- W3129072783 cites W2771486447 @default.
- W3129072783 cites W2783837693 @default.
- W3129072783 cites W2914664131 @default.
- W3129072783 cites W2962916234 @default.
- W3129072783 cites W2994974226 @default.
- W3129072783 cites W3027447579 @default.
- W3129072783 cites W3031875317 @default.
- W3129072783 cites W3047180765 @default.
- W3129072783 cites W3085694362 @default.
- W3129072783 doi "https://doi.org/10.1016/j.patcog.2022.108528" @default.
- W3129072783 hasPublicationYear "2022" @default.
- W3129072783 type Work @default.
- W3129072783 sameAs 3129072783 @default.
- W3129072783 citedByCount "22" @default.
- W3129072783 countsByYear W31290727832021 @default.
- W3129072783 countsByYear W31290727832022 @default.
- W3129072783 countsByYear W31290727832023 @default.
- W3129072783 crossrefType "journal-article" @default.
- W3129072783 hasAuthorship W3129072783A5034095159 @default.
- W3129072783 hasAuthorship W3129072783A5055128383 @default.
- W3129072783 hasAuthorship W3129072783A5057063594 @default.
- W3129072783 hasAuthorship W3129072783A5080301434 @default.
- W3129072783 hasAuthorship W3129072783A5083626840 @default.
- W3129072783 hasBestOaLocation W31290727832 @default.
- W3129072783 hasConcept C108583219 @default.
- W3129072783 hasConcept C11413529 @default.
- W3129072783 hasConcept C115961682 @default.
- W3129072783 hasConcept C119857082 @default.
- W3129072783 hasConcept C134306372 @default.
- W3129072783 hasConcept C153180895 @default.
- W3129072783 hasConcept C154945302 @default.
- W3129072783 hasConcept C165696696 @default.
- W3129072783 hasConcept C1667742 @default.
- W3129072783 hasConcept C207685749 @default.
- W3129072783 hasConcept C2776214188 @default.
- W3129072783 hasConcept C2779231336 @default.
- W3129072783 hasConcept C33923547 @default.
- W3129072783 hasConcept C36503486 @default.
- W3129072783 hasConcept C38652104 @default.
- W3129072783 hasConcept C41008148 @default.
- W3129072783 hasConcept C75294576 @default.
- W3129072783 hasConcept C81363708 @default.
- W3129072783 hasConcept C97931131 @default.
- W3129072783 hasConceptScore W3129072783C108583219 @default.
- W3129072783 hasConceptScore W3129072783C11413529 @default.
- W3129072783 hasConceptScore W3129072783C115961682 @default.
- W3129072783 hasConceptScore W3129072783C119857082 @default.
- W3129072783 hasConceptScore W3129072783C134306372 @default.
- W3129072783 hasConceptScore W3129072783C153180895 @default.
- W3129072783 hasConceptScore W3129072783C154945302 @default.
- W3129072783 hasConceptScore W3129072783C165696696 @default.
- W3129072783 hasConceptScore W3129072783C1667742 @default.
- W3129072783 hasConceptScore W3129072783C207685749 @default.
- W3129072783 hasConceptScore W3129072783C2776214188 @default.
- W3129072783 hasConceptScore W3129072783C2779231336 @default.
- W3129072783 hasConceptScore W3129072783C33923547 @default.
- W3129072783 hasConceptScore W3129072783C36503486 @default.
- W3129072783 hasConceptScore W3129072783C38652104 @default.
- W3129072783 hasConceptScore W3129072783C41008148 @default.
- W3129072783 hasConceptScore W3129072783C75294576 @default.
- W3129072783 hasConceptScore W3129072783C81363708 @default.
- W3129072783 hasConceptScore W3129072783C97931131 @default.
- W3129072783 hasLocation W31290727831 @default.
- W3129072783 hasLocation W31290727832 @default.
- W3129072783 hasLocation W31290727833 @default.
- W3129072783 hasOpenAccess W3129072783 @default.
- W3129072783 hasPrimaryLocation W31290727831 @default.
- W3129072783 hasRelatedWork W1502599625 @default.
- W3129072783 hasRelatedWork W2373478030 @default.
- W3129072783 hasRelatedWork W2378679551 @default.
- W3129072783 hasRelatedWork W2378994405 @default.
- W3129072783 hasRelatedWork W2385974820 @default.
- W3129072783 hasRelatedWork W2761785940 @default.
- W3129072783 hasRelatedWork W2971377935 @default.
- W3129072783 hasRelatedWork W3149739944 @default.
- W3129072783 hasRelatedWork W2185495922 @default.
- W3129072783 hasRelatedWork W2746783359 @default.
- W3129072783 hasVolume "126" @default.
- W3129072783 isParatext "false" @default.
- W3129072783 isRetracted "false" @default.
- W3129072783 magId "3129072783" @default.
- W3129072783 workType "article" @default.