Matches in SemOpenAlex for { <https://semopenalex.org/work/W3201428215> ?p ?o ?g. }
- W3201428215 endingPage "2992" @default.
- W3201428215 startingPage "2978" @default.
- W3201428215 abstract "Nowadays, generative adversarial network (GAN) is making continuous breakthroughs in many machine learning tasks. The popular GANs usually involve computation-intensive deconvolution operations, leading to limited real-time applications. Prior works have brought several accelerators for deconvolution, but all of them suffer from severe problems, such as computation imbalance and large memory requirements. In this article, we first introduce a novel fast transformation algorithm (FTA) for deconvolution computation, which well solves the computation imbalance problem and removes the extra memory requirement for overlapped partial sums. Besides, it can reduce the computation complexity for various types of deconvolutions significantly. Based on FTA, we develop a fast computing core (FCC) and the corresponding computing array so that the deconvolution can be efficiently computed. We next optimize the dataflow and storage scheme to further reuse on-chip memory and improve the computation efficiency. Finally, we present a computation-efficient hardware architecture for GANs and validate it on several GAN benchmarks, such as deep convolutional GAN (DCGAN), energy-based GAN (EBGAN), and Wasserstein GAN (WGAN). The experimental results show that our design can reach 2211 GOPS under 185-MHz working frequency on Intel Stratix 10SX field-programmable gate array (FPGA) board with satisfactory visual results. In brief, the proposed design can achieve more than 2× hardware efficiency improvement over previous designs, and it can reduce the storage requirement drastically." @default.
- W3201428215 created "2021-09-27" @default.
- W3201428215 creator A5047445405 @default.
- W3201428215 creator A5051709914 @default.
- W3201428215 creator A5072609627 @default.
- W3201428215 date "2023-06-01" @default.
- W3201428215 modified "2023-09-27" @default.
- W3201428215 title "FTA-GAN: A Computation-Efficient Accelerator for GANs With Fast Transformation Algorithm" @default.
- W3201428215 cites W1834627138 @default.
- W3201428215 cites W1903029394 @default.
- W3201428215 cites W2035856475 @default.
- W3201428215 cites W2141200610 @default.
- W3201428215 cites W2172654076 @default.
- W3201428215 cites W2476548250 @default.
- W3201428215 cites W2503339013 @default.
- W3201428215 cites W2597692291 @default.
- W3201428215 cites W2729080111 @default.
- W3201428215 cites W2768993447 @default.
- W3201428215 cites W2798838583 @default.
- W3201428215 cites W2895738878 @default.
- W3201428215 cites W2899824860 @default.
- W3201428215 cites W2906403229 @default.
- W3201428215 cites W2928560789 @default.
- W3201428215 cites W2940399336 @default.
- W3201428215 cites W2942898010 @default.
- W3201428215 cites W2943690741 @default.
- W3201428215 cites W2962793481 @default.
- W3201428215 cites W2997109118 @default.
- W3201428215 cites W3005283024 @default.
- W3201428215 cites W3005924570 @default.
- W3201428215 cites W3013525232 @default.
- W3201428215 cites W3035072672 @default.
- W3201428215 cites W3040533934 @default.
- W3201428215 cites W3047462917 @default.
- W3201428215 cites W3159063014 @default.
- W3201428215 cites W4243682116 @default.
- W3201428215 cites W4250908627 @default.
- W3201428215 doi "https://doi.org/10.1109/tnnls.2021.3110728" @default.
- W3201428215 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/34534090" @default.
- W3201428215 hasPublicationYear "2023" @default.
- W3201428215 type Work @default.
- W3201428215 sameAs 3201428215 @default.
- W3201428215 citedByCount "4" @default.
- W3201428215 countsByYear W32014282152022 @default.
- W3201428215 crossrefType "journal-article" @default.
- W3201428215 hasAuthorship W3201428215A5047445405 @default.
- W3201428215 hasAuthorship W3201428215A5051709914 @default.
- W3201428215 hasAuthorship W3201428215A5072609627 @default.
- W3201428215 hasConcept C104317684 @default.
- W3201428215 hasConcept C113775141 @default.
- W3201428215 hasConcept C11413529 @default.
- W3201428215 hasConcept C154945302 @default.
- W3201428215 hasConcept C173608175 @default.
- W3201428215 hasConcept C174576160 @default.
- W3201428215 hasConcept C185592680 @default.
- W3201428215 hasConcept C204241405 @default.
- W3201428215 hasConcept C2776277307 @default.
- W3201428215 hasConcept C41008148 @default.
- W3201428215 hasConcept C42935608 @default.
- W3201428215 hasConcept C45347329 @default.
- W3201428215 hasConcept C45374587 @default.
- W3201428215 hasConcept C459310 @default.
- W3201428215 hasConcept C50644808 @default.
- W3201428215 hasConcept C55493867 @default.
- W3201428215 hasConcept C9390403 @default.
- W3201428215 hasConcept C96324660 @default.
- W3201428215 hasConceptScore W3201428215C104317684 @default.
- W3201428215 hasConceptScore W3201428215C113775141 @default.
- W3201428215 hasConceptScore W3201428215C11413529 @default.
- W3201428215 hasConceptScore W3201428215C154945302 @default.
- W3201428215 hasConceptScore W3201428215C173608175 @default.
- W3201428215 hasConceptScore W3201428215C174576160 @default.
- W3201428215 hasConceptScore W3201428215C185592680 @default.
- W3201428215 hasConceptScore W3201428215C204241405 @default.
- W3201428215 hasConceptScore W3201428215C2776277307 @default.
- W3201428215 hasConceptScore W3201428215C41008148 @default.
- W3201428215 hasConceptScore W3201428215C42935608 @default.
- W3201428215 hasConceptScore W3201428215C45347329 @default.
- W3201428215 hasConceptScore W3201428215C45374587 @default.
- W3201428215 hasConceptScore W3201428215C459310 @default.
- W3201428215 hasConceptScore W3201428215C50644808 @default.
- W3201428215 hasConceptScore W3201428215C55493867 @default.
- W3201428215 hasConceptScore W3201428215C9390403 @default.
- W3201428215 hasConceptScore W3201428215C96324660 @default.
- W3201428215 hasFunder F4320321001 @default.
- W3201428215 hasFunder F4320335787 @default.
- W3201428215 hasIssue "6" @default.
- W3201428215 hasLocation W32014282151 @default.
- W3201428215 hasLocation W32014282152 @default.
- W3201428215 hasOpenAccess W3201428215 @default.
- W3201428215 hasPrimaryLocation W32014282151 @default.
- W3201428215 hasRelatedWork W1535188164 @default.
- W3201428215 hasRelatedWork W1572523360 @default.
- W3201428215 hasRelatedWork W1601492201 @default.
- W3201428215 hasRelatedWork W2047588290 @default.
- W3201428215 hasRelatedWork W2068231627 @default.
- W3201428215 hasRelatedWork W2371372853 @default.
- W3201428215 hasRelatedWork W3103072756 @default.