Matches in SemOpenAlex for { <https://semopenalex.org/work/W3193148745> ?p ?o ?g. }
- W3193148745 abstract "Along with the rapid development of real-world applications, higher requirements on the accuracy and efficiency of image super-resolution (SR) are brought forward. Though existing methods have achieved remarkable success, the majority of them demand plenty of computational resources and large amount of RAM, and thus they can not be well applied to mobile device. In this paper, we aim at designing efficient architecture for 8-bit quantization and deploy it on mobile device. First, we conduct an experiment about meta-node latency by decomposing lightweight SR architectures, which determines the portable operations we can utilize. Then, we dig deeper into what kind of architecture is beneficial to 8-bit quantization and propose anchor-based plain net (ABPN). Finally, we adopt quantization-aware training strategy to further boost the performance. Our model can outperform 8-bit quantized FSRCNN by nearly 2dB in terms of PSNR, while satisfying realistic needs at the same time. Code is avaliable at https://github.com/NJU-Jet/SR_Mobile_Quantization." @default.
- W3193148745 created "2021-08-16" @default.
- W3193148745 creator A5021809690 @default.
- W3193148745 creator A5085352453 @default.
- W3193148745 creator A5089631680 @default.
- W3193148745 creator A5089775519 @default.
- W3193148745 date "2021-06-01" @default.
- W3193148745 modified "2023-10-18" @default.
- W3193148745 title "Anchor-based Plain Net for Mobile Image Super-Resolution" @default.
- W3193148745 cites W1677182931 @default.
- W3193148745 cites W1949096787 @default.
- W3193148745 cites W2010133640 @default.
- W3193148745 cites W2047264196 @default.
- W3193148745 cites W2051738325 @default.
- W3193148745 cites W2058182006 @default.
- W3193148745 cites W2097074225 @default.
- W3193148745 cites W2103844245 @default.
- W3193148745 cites W2118963448 @default.
- W3193148745 cites W2137987353 @default.
- W3193148745 cites W2146782367 @default.
- W3193148745 cites W2157190232 @default.
- W3193148745 cites W2183341477 @default.
- W3193148745 cites W2194775991 @default.
- W3193148745 cites W2214802144 @default.
- W3193148745 cites W2242218935 @default.
- W3193148745 cites W2476548250 @default.
- W3193148745 cites W2607041014 @default.
- W3193148745 cites W2739757502 @default.
- W3193148745 cites W2747898905 @default.
- W3193148745 cites W2781674358 @default.
- W3193148745 cites W2795024892 @default.
- W3193148745 cites W2866634454 @default.
- W3193148745 cites W2954930822 @default.
- W3193148745 cites W2963037581 @default.
- W3193148745 cites W2963372104 @default.
- W3193148745 cites W2963470893 @default.
- W3193148745 cites W2963645458 @default.
- W3193148745 cites W2964101377 @default.
- W3193148745 cites W2964125708 @default.
- W3193148745 cites W2986556233 @default.
- W3193148745 cites W2998506323 @default.
- W3193148745 cites W3011005573 @default.
- W3193148745 cites W3034439835 @default.
- W3193148745 cites W3043571714 @default.
- W3193148745 cites W3105328221 @default.
- W3193148745 cites W3110378228 @default.
- W3193148745 doi "https://doi.org/10.1109/cvprw53098.2021.00283" @default.
- W3193148745 hasPublicationYear "2021" @default.
- W3193148745 type Work @default.
- W3193148745 sameAs 3193148745 @default.
- W3193148745 citedByCount "27" @default.
- W3193148745 countsByYear W31931487452022 @default.
- W3193148745 countsByYear W31931487452023 @default.
- W3193148745 crossrefType "proceedings-article" @default.
- W3193148745 hasAuthorship W3193148745A5021809690 @default.
- W3193148745 hasAuthorship W3193148745A5085352453 @default.
- W3193148745 hasAuthorship W3193148745A5089631680 @default.
- W3193148745 hasAuthorship W3193148745A5089775519 @default.
- W3193148745 hasBestOaLocation W31931487452 @default.
- W3193148745 hasConcept C111919701 @default.
- W3193148745 hasConcept C113775141 @default.
- W3193148745 hasConcept C11413529 @default.
- W3193148745 hasConcept C118524514 @default.
- W3193148745 hasConcept C123657996 @default.
- W3193148745 hasConcept C142362112 @default.
- W3193148745 hasConcept C153349607 @default.
- W3193148745 hasConcept C177264268 @default.
- W3193148745 hasConcept C186967261 @default.
- W3193148745 hasConcept C199360897 @default.
- W3193148745 hasConcept C2776760102 @default.
- W3193148745 hasConcept C28855332 @default.
- W3193148745 hasConcept C41008148 @default.
- W3193148745 hasConcept C76155785 @default.
- W3193148745 hasConcept C79403827 @default.
- W3193148745 hasConcept C82876162 @default.
- W3193148745 hasConceptScore W3193148745C111919701 @default.
- W3193148745 hasConceptScore W3193148745C113775141 @default.
- W3193148745 hasConceptScore W3193148745C11413529 @default.
- W3193148745 hasConceptScore W3193148745C118524514 @default.
- W3193148745 hasConceptScore W3193148745C123657996 @default.
- W3193148745 hasConceptScore W3193148745C142362112 @default.
- W3193148745 hasConceptScore W3193148745C153349607 @default.
- W3193148745 hasConceptScore W3193148745C177264268 @default.
- W3193148745 hasConceptScore W3193148745C186967261 @default.
- W3193148745 hasConceptScore W3193148745C199360897 @default.
- W3193148745 hasConceptScore W3193148745C2776760102 @default.
- W3193148745 hasConceptScore W3193148745C28855332 @default.
- W3193148745 hasConceptScore W3193148745C41008148 @default.
- W3193148745 hasConceptScore W3193148745C76155785 @default.
- W3193148745 hasConceptScore W3193148745C79403827 @default.
- W3193148745 hasConceptScore W3193148745C82876162 @default.
- W3193148745 hasLocation W31931487451 @default.
- W3193148745 hasLocation W31931487452 @default.
- W3193148745 hasOpenAccess W3193148745 @default.
- W3193148745 hasPrimaryLocation W31931487451 @default.
- W3193148745 hasRelatedWork W1987753576 @default.
- W3193148745 hasRelatedWork W2052158784 @default.
- W3193148745 hasRelatedWork W2373724792 @default.
- W3193148745 hasRelatedWork W2380362089 @default.
- W3193148745 hasRelatedWork W3033499831 @default.