Matches in SemOpenAlex for { <https://semopenalex.org/work/W4379115860> ?p ?o ?g. }
Showing items 1 to 71 of
71
with 100 items per page.
- W4379115860 abstract "Few-shot learning (FSL) is a promising meta-learning paradigm that trains classification models on the fly with a few training samples. However, existing FSL classifiers are either computationally expensive, or are not accurate enough. In this work, we propose an efficient in-memory FSL classifier, FSL-HD, based on hyperdimensional computing (HDC) that achieves state-of-the-art FSL accuracy and efficiency. We devise an HDC-based FSL framework with efficient HDC encoding and search to reduce high complexity caused by the large dimensionality. Also, we design a scalable in-memory architecture to accelerate FSL-HD on ReRAM with distributed dataflow and organization that maximizes the data parallelism and hardware utilization. The evaluation shows that FSL-HD achieves 4.2% higher accuracy compared to other FSL classifiers. FSL-HD achieves <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$100-1000times$</tex> better energy efficiency and <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$9-66times$</tex> speedup over the CPU and GPU baselines. Moreover, FSL-HD is more accurate, scalable and <tex xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>$2.5times$</tex> faster than the state-of-the-art ReRAM-based FSL design, SAPIENS, while requiring 85% less area." @default.
- W4379115860 created "2023-06-03" @default.
- W4379115860 creator A5025573294 @default.
- W4379115860 creator A5039571679 @default.
- W4379115860 creator A5072579616 @default.
- W4379115860 date "2023-04-01" @default.
- W4379115860 modified "2023-09-23" @default.
- W4379115860 title "FSL-HD: Accelerating Few-Shot Learning on ReRAM using Hyperdimensional Computing" @default.
- W4379115860 cites W2002427601 @default.
- W4379115860 cites W2047643928 @default.
- W4379115860 cites W2194321275 @default.
- W4379115860 cites W2194775991 @default.
- W4379115860 cites W2197984537 @default.
- W4379115860 cites W3091977376 @default.
- W4379115860 cites W3108975329 @default.
- W4379115860 cites W3120093386 @default.
- W4379115860 cites W3200825586 @default.
- W4379115860 doi "https://doi.org/10.23919/date56975.2023.10136901" @default.
- W4379115860 hasPublicationYear "2023" @default.
- W4379115860 type Work @default.
- W4379115860 citedByCount "0" @default.
- W4379115860 crossrefType "proceedings-article" @default.
- W4379115860 hasAuthorship W4379115860A5025573294 @default.
- W4379115860 hasAuthorship W4379115860A5039571679 @default.
- W4379115860 hasAuthorship W4379115860A5072579616 @default.
- W4379115860 hasConcept C111919701 @default.
- W4379115860 hasConcept C119857082 @default.
- W4379115860 hasConcept C147789679 @default.
- W4379115860 hasConcept C149635348 @default.
- W4379115860 hasConcept C154945302 @default.
- W4379115860 hasConcept C173608175 @default.
- W4379115860 hasConcept C17525397 @default.
- W4379115860 hasConcept C182019814 @default.
- W4379115860 hasConcept C185592680 @default.
- W4379115860 hasConcept C2780513914 @default.
- W4379115860 hasConcept C41008148 @default.
- W4379115860 hasConcept C48044578 @default.
- W4379115860 hasConcept C68339613 @default.
- W4379115860 hasConcept C95623464 @default.
- W4379115860 hasConcept C96324660 @default.
- W4379115860 hasConceptScore W4379115860C111919701 @default.
- W4379115860 hasConceptScore W4379115860C119857082 @default.
- W4379115860 hasConceptScore W4379115860C147789679 @default.
- W4379115860 hasConceptScore W4379115860C149635348 @default.
- W4379115860 hasConceptScore W4379115860C154945302 @default.
- W4379115860 hasConceptScore W4379115860C173608175 @default.
- W4379115860 hasConceptScore W4379115860C17525397 @default.
- W4379115860 hasConceptScore W4379115860C182019814 @default.
- W4379115860 hasConceptScore W4379115860C185592680 @default.
- W4379115860 hasConceptScore W4379115860C2780513914 @default.
- W4379115860 hasConceptScore W4379115860C41008148 @default.
- W4379115860 hasConceptScore W4379115860C48044578 @default.
- W4379115860 hasConceptScore W4379115860C68339613 @default.
- W4379115860 hasConceptScore W4379115860C95623464 @default.
- W4379115860 hasConceptScore W4379115860C96324660 @default.
- W4379115860 hasLocation W43791158601 @default.
- W4379115860 hasOpenAccess W4379115860 @default.
- W4379115860 hasPrimaryLocation W43791158601 @default.
- W4379115860 hasRelatedWork W184060744 @default.
- W4379115860 hasRelatedWork W2100229967 @default.
- W4379115860 hasRelatedWork W2134266179 @default.
- W4379115860 hasRelatedWork W2370911386 @default.
- W4379115860 hasRelatedWork W2470589840 @default.
- W4379115860 hasRelatedWork W3013306997 @default.
- W4379115860 hasRelatedWork W3120511008 @default.
- W4379115860 hasRelatedWork W3207233976 @default.
- W4379115860 hasRelatedWork W4239886135 @default.
- W4379115860 hasRelatedWork W4286901967 @default.
- W4379115860 isParatext "false" @default.
- W4379115860 isRetracted "false" @default.
- W4379115860 workType "article" @default.