Matches in SemOpenAlex for { <https://semopenalex.org/work/W3011010299> ?p ?o ?g. }
Showing items 1 to 74 of
74
with 100 items per page.
- W3011010299 abstract "The recurrent neural networks (RNNs) along with connectionist temporal classification (CTC) have been widely used in many sequence to sequence tasks, including automatic speech recognition (ASR), lipreading, and scene text recognition (STR). In these systems, CTC-trained RNNs usually require specific CTC-decoders after their output layers. Many existing CTC-trained RNN inference systems use FPGA to do calculations of RNNs, and decode their outputs on CPU. However, with the development of FPGA-based RNN hardware accelerators, existing CPU-based CTC-decoder can not meet the latency requirement of them. To resolve this issue, this paper proposes an efficient hardware architecture for the CTC beam search decoder based on the decoding method reported in our previous work. The experimental results show that the system latency per sample of the CTC-decoder is only 7.19us on Xilinx xc7vx1140tflg19301 FPGA platform, which is lower than state-of-the-art RNNs. We also implement the origin algorithm on the same FPGA platform. Comparison results show that the improved one reduces the system latency per sample by 63.67%, the LUTRAMs by 97.44%, the FFs by 79.55%, and the DSPs by 50%. To the best of our knowledge, this is the first work on hardware implementation for CTC beam search decoder." @default.
- W3011010299 created "2020-03-23" @default.
- W3011010299 creator A5019940239 @default.
- W3011010299 creator A5039665206 @default.
- W3011010299 creator A5047445405 @default.
- W3011010299 creator A5059107361 @default.
- W3011010299 creator A5084117570 @default.
- W3011010299 date "2019-10-01" @default.
- W3011010299 modified "2023-09-23" @default.
- W3011010299 title "A Low-Latency and Low-Complexity Hardware Architecture for CTC Beam Search Decoding" @default.
- W3011010299 cites W2127141656 @default.
- W3011010299 cites W2144499799 @default.
- W3011010299 cites W2194187530 @default.
- W3011010299 cites W2585720638 @default.
- W3011010299 cites W2608093348 @default.
- W3011010299 cites W2910396952 @default.
- W3011010299 cites W2911291251 @default.
- W3011010299 cites W2940345045 @default.
- W3011010299 cites W2944322787 @default.
- W3011010299 cites W2962820060 @default.
- W3011010299 cites W2962962542 @default.
- W3011010299 cites W2963030892 @default.
- W3011010299 cites W2963537349 @default.
- W3011010299 doi "https://doi.org/10.1109/sips47522.2019.9020324" @default.
- W3011010299 hasPublicationYear "2019" @default.
- W3011010299 type Work @default.
- W3011010299 sameAs 3011010299 @default.
- W3011010299 citedByCount "1" @default.
- W3011010299 countsByYear W30110102992022 @default.
- W3011010299 crossrefType "proceedings-article" @default.
- W3011010299 hasAuthorship W3011010299A5019940239 @default.
- W3011010299 hasAuthorship W3011010299A5039665206 @default.
- W3011010299 hasAuthorship W3011010299A5047445405 @default.
- W3011010299 hasAuthorship W3011010299A5059107361 @default.
- W3011010299 hasAuthorship W3011010299A5084117570 @default.
- W3011010299 hasConcept C118524514 @default.
- W3011010299 hasConcept C123657996 @default.
- W3011010299 hasConcept C142362112 @default.
- W3011010299 hasConcept C153349607 @default.
- W3011010299 hasConcept C31258907 @default.
- W3011010299 hasConcept C41008148 @default.
- W3011010299 hasConcept C46637626 @default.
- W3011010299 hasConcept C57273362 @default.
- W3011010299 hasConcept C76155785 @default.
- W3011010299 hasConcept C82876162 @default.
- W3011010299 hasConcept C9390403 @default.
- W3011010299 hasConceptScore W3011010299C118524514 @default.
- W3011010299 hasConceptScore W3011010299C123657996 @default.
- W3011010299 hasConceptScore W3011010299C142362112 @default.
- W3011010299 hasConceptScore W3011010299C153349607 @default.
- W3011010299 hasConceptScore W3011010299C31258907 @default.
- W3011010299 hasConceptScore W3011010299C41008148 @default.
- W3011010299 hasConceptScore W3011010299C46637626 @default.
- W3011010299 hasConceptScore W3011010299C57273362 @default.
- W3011010299 hasConceptScore W3011010299C76155785 @default.
- W3011010299 hasConceptScore W3011010299C82876162 @default.
- W3011010299 hasConceptScore W3011010299C9390403 @default.
- W3011010299 hasLocation W30110102991 @default.
- W3011010299 hasOpenAccess W3011010299 @default.
- W3011010299 hasPrimaryLocation W30110102991 @default.
- W3011010299 hasRelatedWork W1833747146 @default.
- W3011010299 hasRelatedWork W1974303716 @default.
- W3011010299 hasRelatedWork W2226826524 @default.
- W3011010299 hasRelatedWork W2367936931 @default.
- W3011010299 hasRelatedWork W2383828164 @default.
- W3011010299 hasRelatedWork W2909132900 @default.
- W3011010299 hasRelatedWork W3196187157 @default.
- W3011010299 hasRelatedWork W3205411230 @default.
- W3011010299 hasRelatedWork W396164270 @default.
- W3011010299 hasRelatedWork W4286899009 @default.
- W3011010299 isParatext "false" @default.
- W3011010299 isRetracted "false" @default.
- W3011010299 magId "3011010299" @default.
- W3011010299 workType "article" @default.