Matches in SemOpenAlex for { <https://semopenalex.org/work/W3183327882> ?p ?o ?g. }
- W3183327882 abstract "Abstract In this work a novel architecture, named pseudo-softmax, to compute an approximated form of the softmax function is presented. This architecture can be fruitfully used in the last layer of Neural Networks and Convolutional Neural Networks for classification tasks, and in Reinforcement Learning hardware accelerators to compute the Boltzmann action-selection policy. The proposed pseudo-softmax design, intended for efficient hardware implementation, exploits the typical integer quantization of hardware-based Neural Networks obtaining an accurate approximation of the result. In the paper, a detailed description of the architecture is given and an extensive analysis of the approximation error is performed by using both custom stimuli and real-world Convolutional Neural Networks inputs. The implementation results, based on CMOS standard-cell technology, compared to state-of-the-art architectures show reduced approximation errors." @default.
- W3183327882 created "2021-08-02" @default.
- W3183327882 creator A5000963583 @default.
- W3183327882 creator A5022318517 @default.
- W3183327882 creator A5028297120 @default.
- W3183327882 creator A5044774713 @default.
- W3183327882 creator A5048482308 @default.
- W3183327882 creator A5063300969 @default.
- W3183327882 creator A5075532616 @default.
- W3183327882 date "2021-07-28" @default.
- W3183327882 modified "2023-10-18" @default.
- W3183327882 title "A pseudo-softmax function for hardware-based high speed image classification" @default.
- W3183327882 cites W2104226279 @default.
- W3183327882 cites W2117539524 @default.
- W3183327882 cites W2157348988 @default.
- W3183327882 cites W2183341477 @default.
- W3183327882 cites W2194775991 @default.
- W3183327882 cites W2608093348 @default.
- W3183327882 cites W2891483719 @default.
- W3183327882 cites W2903640996 @default.
- W3183327882 cites W2910396952 @default.
- W3183327882 cites W2914968962 @default.
- W3183327882 cites W2937064326 @default.
- W3183327882 cites W2946480111 @default.
- W3183327882 cites W2951045917 @default.
- W3183327882 cites W2963163009 @default.
- W3183327882 cites W2965972813 @default.
- W3183327882 cites W2997106510 @default.
- W3183327882 cites W2998165854 @default.
- W3183327882 cites W3038988173 @default.
- W3183327882 cites W3082834449 @default.
- W3183327882 cites W3094121892 @default.
- W3183327882 cites W3125827992 @default.
- W3183327882 cites W4249932213 @default.
- W3183327882 cites W2906917284 @default.
- W3183327882 doi "https://doi.org/10.1038/s41598-021-94691-7" @default.
- W3183327882 hasPubMedCentralId "https://www.ncbi.nlm.nih.gov/pmc/articles/8319144" @default.
- W3183327882 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/34321514" @default.
- W3183327882 hasPublicationYear "2021" @default.
- W3183327882 type Work @default.
- W3183327882 sameAs 3183327882 @default.
- W3183327882 citedByCount "22" @default.
- W3183327882 countsByYear W31833278822021 @default.
- W3183327882 countsByYear W31833278822022 @default.
- W3183327882 countsByYear W31833278822023 @default.
- W3183327882 crossrefType "journal-article" @default.
- W3183327882 hasAuthorship W3183327882A5000963583 @default.
- W3183327882 hasAuthorship W3183327882A5022318517 @default.
- W3183327882 hasAuthorship W3183327882A5028297120 @default.
- W3183327882 hasAuthorship W3183327882A5044774713 @default.
- W3183327882 hasAuthorship W3183327882A5048482308 @default.
- W3183327882 hasAuthorship W3183327882A5063300969 @default.
- W3183327882 hasAuthorship W3183327882A5075532616 @default.
- W3183327882 hasBestOaLocation W31833278821 @default.
- W3183327882 hasConcept C108583219 @default.
- W3183327882 hasConcept C113775141 @default.
- W3183327882 hasConcept C11413529 @default.
- W3183327882 hasConcept C115961682 @default.
- W3183327882 hasConcept C118524514 @default.
- W3183327882 hasConcept C123657996 @default.
- W3183327882 hasConcept C142362112 @default.
- W3183327882 hasConcept C153349607 @default.
- W3183327882 hasConcept C154945302 @default.
- W3183327882 hasConcept C173608175 @default.
- W3183327882 hasConcept C188441871 @default.
- W3183327882 hasConcept C199360897 @default.
- W3183327882 hasConcept C2777904410 @default.
- W3183327882 hasConcept C28855332 @default.
- W3183327882 hasConcept C41008148 @default.
- W3183327882 hasConcept C50644808 @default.
- W3183327882 hasConcept C65232700 @default.
- W3183327882 hasConcept C68339613 @default.
- W3183327882 hasConcept C75294576 @default.
- W3183327882 hasConcept C81363708 @default.
- W3183327882 hasConceptScore W3183327882C108583219 @default.
- W3183327882 hasConceptScore W3183327882C113775141 @default.
- W3183327882 hasConceptScore W3183327882C11413529 @default.
- W3183327882 hasConceptScore W3183327882C115961682 @default.
- W3183327882 hasConceptScore W3183327882C118524514 @default.
- W3183327882 hasConceptScore W3183327882C123657996 @default.
- W3183327882 hasConceptScore W3183327882C142362112 @default.
- W3183327882 hasConceptScore W3183327882C153349607 @default.
- W3183327882 hasConceptScore W3183327882C154945302 @default.
- W3183327882 hasConceptScore W3183327882C173608175 @default.
- W3183327882 hasConceptScore W3183327882C188441871 @default.
- W3183327882 hasConceptScore W3183327882C199360897 @default.
- W3183327882 hasConceptScore W3183327882C2777904410 @default.
- W3183327882 hasConceptScore W3183327882C28855332 @default.
- W3183327882 hasConceptScore W3183327882C41008148 @default.
- W3183327882 hasConceptScore W3183327882C50644808 @default.
- W3183327882 hasConceptScore W3183327882C65232700 @default.
- W3183327882 hasConceptScore W3183327882C68339613 @default.
- W3183327882 hasConceptScore W3183327882C75294576 @default.
- W3183327882 hasConceptScore W3183327882C81363708 @default.
- W3183327882 hasIssue "1" @default.
- W3183327882 hasLocation W31833278821 @default.
- W3183327882 hasLocation W31833278822 @default.
- W3183327882 hasLocation W31833278823 @default.
- W3183327882 hasLocation W31833278824 @default.
- W3183327882 hasLocation W31833278825 @default.