Matches in SemOpenAlex for { <https://semopenalex.org/work/W4289754489> ?p ?o ?g. }
Showing items 1 to 76 of
76
with 100 items per page.
- W4289754489 abstract "It is well known that many types of artificial neural networks, including recurrent networks, can achieve a high classification accuracy even with low-precision weights and activations. The reduction in precision generally yields much more efficient hardware implementations in regards to hardware cost, memory requirements, energy, and achievable throughput. In this paper, we present the first systematic exploration of this design space as a function of precision for Bidirectional Long Short-Term Memory (BiLSTM) neural network. Specifically, we include an in-depth investigation of precision vs. accuracy using a fully hardware-aware training flow, where during training quantization of all aspects of the network including weights, input, output and in-memory cell activations are taken into consideration. In addition, hardware resource cost, power consumption and throughput scalability are explored as a function of precision for FPGA-based implementations of BiLSTM, and multiple approaches of parallelizing the hardware. We provide the first open source HLS library extension of FINN for parameterizable hardware architectures of LSTM layers on FPGAs which offers full precision flexibility and allows for parameterizable performance scaling offering different levels of parallelism within the architecture. Based on this library, we present an FPGA-based accelerator for BiLSTM neural network designed for optical character recognition, along with numerous other experimental proof points for a Zynq UltraScale+ XCZU7EV MPSoC within the given design space." @default.
- W4289754489 created "2022-08-04" @default.
- W4289754489 creator A5026761218 @default.
- W4289754489 creator A5031107943 @default.
- W4289754489 creator A5031192719 @default.
- W4289754489 creator A5057376276 @default.
- W4289754489 creator A5059285190 @default.
- W4289754489 creator A5076669165 @default.
- W4289754489 date "2018-07-11" @default.
- W4289754489 modified "2023-09-25" @default.
- W4289754489 title "FINN-L: Library Extensions and Design Trade-off Analysis for Variable Precision LSTM Networks on FPGAs" @default.
- W4289754489 doi "https://doi.org/10.48550/arxiv.1807.04093" @default.
- W4289754489 hasPublicationYear "2018" @default.
- W4289754489 type Work @default.
- W4289754489 citedByCount "0" @default.
- W4289754489 crossrefType "posted-content" @default.
- W4289754489 hasAuthorship W4289754489A5026761218 @default.
- W4289754489 hasAuthorship W4289754489A5031107943 @default.
- W4289754489 hasAuthorship W4289754489A5031192719 @default.
- W4289754489 hasAuthorship W4289754489A5057376276 @default.
- W4289754489 hasAuthorship W4289754489A5059285190 @default.
- W4289754489 hasAuthorship W4289754489A5076669165 @default.
- W4289754489 hasBestOaLocation W42897544891 @default.
- W4289754489 hasConcept C113775141 @default.
- W4289754489 hasConcept C118021083 @default.
- W4289754489 hasConcept C118524514 @default.
- W4289754489 hasConcept C149635348 @default.
- W4289754489 hasConcept C154945302 @default.
- W4289754489 hasConcept C157764524 @default.
- W4289754489 hasConcept C199360897 @default.
- W4289754489 hasConcept C26713055 @default.
- W4289754489 hasConcept C2776221188 @default.
- W4289754489 hasConcept C2777187653 @default.
- W4289754489 hasConcept C41008148 @default.
- W4289754489 hasConcept C42935608 @default.
- W4289754489 hasConcept C48044578 @default.
- W4289754489 hasConcept C50644808 @default.
- W4289754489 hasConcept C555944384 @default.
- W4289754489 hasConcept C76155785 @default.
- W4289754489 hasConcept C77088390 @default.
- W4289754489 hasConcept C9390403 @default.
- W4289754489 hasConceptScore W4289754489C113775141 @default.
- W4289754489 hasConceptScore W4289754489C118021083 @default.
- W4289754489 hasConceptScore W4289754489C118524514 @default.
- W4289754489 hasConceptScore W4289754489C149635348 @default.
- W4289754489 hasConceptScore W4289754489C154945302 @default.
- W4289754489 hasConceptScore W4289754489C157764524 @default.
- W4289754489 hasConceptScore W4289754489C199360897 @default.
- W4289754489 hasConceptScore W4289754489C26713055 @default.
- W4289754489 hasConceptScore W4289754489C2776221188 @default.
- W4289754489 hasConceptScore W4289754489C2777187653 @default.
- W4289754489 hasConceptScore W4289754489C41008148 @default.
- W4289754489 hasConceptScore W4289754489C42935608 @default.
- W4289754489 hasConceptScore W4289754489C48044578 @default.
- W4289754489 hasConceptScore W4289754489C50644808 @default.
- W4289754489 hasConceptScore W4289754489C555944384 @default.
- W4289754489 hasConceptScore W4289754489C76155785 @default.
- W4289754489 hasConceptScore W4289754489C77088390 @default.
- W4289754489 hasConceptScore W4289754489C9390403 @default.
- W4289754489 hasLocation W42897544891 @default.
- W4289754489 hasLocation W42897544892 @default.
- W4289754489 hasOpenAccess W4289754489 @default.
- W4289754489 hasPrimaryLocation W42897544891 @default.
- W4289754489 hasRelatedWork W142451177 @default.
- W4289754489 hasRelatedWork W1732210391 @default.
- W4289754489 hasRelatedWork W1975010174 @default.
- W4289754489 hasRelatedWork W2020034302 @default.
- W4289754489 hasRelatedWork W2092276663 @default.
- W4289754489 hasRelatedWork W2149777400 @default.
- W4289754489 hasRelatedWork W2547383453 @default.
- W4289754489 hasRelatedWork W3147787617 @default.
- W4289754489 hasRelatedWork W4297990507 @default.
- W4289754489 hasRelatedWork W630367077 @default.
- W4289754489 isParatext "false" @default.
- W4289754489 isRetracted "false" @default.
- W4289754489 workType "article" @default.