FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4384694832> ?p ?o ?g. }

• W4384694832 abstract "We increasingly rely on deep learning algorithms to process colossal amount of unstructured visual data. Commonly, these deep learning algorithms are deployed as software models on digital hardware, predominantly in data centers. Intrinsic high energy consumption of Cloud-based deployment of deep neural networks (DNNs) inspired researchers to look for alternatives, resulting in a high interest in Spiking Neural Networks (SNNs) and dedicated mixed-signal neuromorphic hardware. As a result, there is an emerging challenge to transfer DNN architecture functionality to energy-efficient spiking non-volatile memory (NVM)-based hardware with minimal loss in the accuracy of visual data processing. Convolutional Neural Network (CNN) is the staple choice of DNN for visual data processing. However, the lack of analog-friendly spiking implementations and alternatives for some core CNN functions, such as MaxPool, hinders the conversion of CNNs into the spike domain, thus hampering neuromorphic hardware development. To address this gap, in this work, we propose MaxPool with temporal multiplexing for Spiking CNNs (SCNNs), which is amenable for implementation in mixed-signal circuits. In this work, we leverage the temporal dynamics of internal membrane potential of Integrate & Fire neurons to enable MaxPool decision-making in the spiking domain. The proposed MaxPool models are implemented and tested within the SCNN architecture using a modified version of the aihwkit framework, a PyTorch-based toolkit for modeling and simulating hardware-based neural networks. The proposed spiking MaxPool scheme can decide even before the complete spatiotemporal input is applied, thus selectively trading off latency with accuracy. It is observed that by allocating just 10% of the spatiotemporal input window for a pooling decision, the proposed spiking MaxPool achieves up to 61.74% accuracy with a 2-bit weight resolution in the CIFAR10 dataset classification task after training with back propagation, with only about 1% performance drop compared to 62.78% accuracy of the 100% spatiotemporal window case with the 2-bit weight resolution to reflect foundry-integrated ReRAM limitations. In addition, we propose the realization of one of the proposed spiking MaxPool techniques in an NVM crossbar array along with periphery circuits designed in a 130nm CMOS technology. The energy-efficiency estimation results show competitive performance compared to recent neuromorphic chip designs." @default.
• W4384694832 created "2023-07-20" @default.
• W4384694832 creator A5028809607 @default.
• W4384694832 creator A5071014444 @default.
• W4384694832 date "2023-07-18" @default.
• W4384694832 modified "2023-09-25" @default.
• W4384694832 title "Spiking CMOS-NVM mixed-signal neuromorphic ConvNet with circuit- and training-optimized temporal subsampling" @default.
• W4384694832 cites W1935332602 @default.
• W4384694832 cites W1944085573 @default.
• W4384694832 cites W2007815184 @default.
• W4384694832 cites W2015548022 @default.
• W4384694832 cites W2020096355 @default.
• W4384694832 cites W2098881438 @default.
• W4384694832 cites W2109596721 @default.
• W4384694832 cites W2122640006 @default.
• W4384694832 cites W2147800946 @default.
• W4384694832 cites W2314470091 @default.
• W4384694832 cites W2556583623 @default.
• W4384694832 cites W2561926108 @default.
• W4384694832 cites W2565565355 @default.
• W4384694832 cites W2621826044 @default.
• W4384694832 cites W2742439472 @default.
• W4384694832 cites W2765741254 @default.
• W4384694832 cites W2775079417 @default.
• W4384694832 cites W2784025056 @default.
• W4384694832 cites W2798808759 @default.
• W4384694832 cites W2805087868 @default.
• W4384694832 cites W2805362231 @default.
• W4384694832 cites W2889463071 @default.
• W4384694832 cites W2897623300 @default.
• W4384694832 cites W2907700607 @default.
• W4384694832 cites W2922002199 @default.
• W4384694832 cites W2947861270 @default.
• W4384694832 cites W2962734882 @default.
• W4384694832 cites W2963150511 @default.
• W4384694832 cites W2964338223 @default.
• W4384694832 cites W2969812992 @default.
• W4384694832 cites W2992817139 @default.
• W4384694832 cites W3006299183 @default.
• W4384694832 cites W3013080934 @default.
• W4384694832 cites W3015982917 @default.
• W4384694832 cites W3080976800 @default.
• W4384694832 cites W3091265339 @default.
• W4384694832 cites W3091740059 @default.
• W4384694832 cites W3095988771 @default.
• W4384694832 cites W3112331435 @default.
• W4384694832 cites W3115618622 @default.
• W4384694832 cites W3124546450 @default.
• W4384694832 cites W3126544799 @default.
• W4384694832 cites W3133754064 @default.
• W4384694832 cites W3134703406 @default.
• W4384694832 cites W3134784623 @default.
• W4384694832 cites W3135783778 @default.
• W4384694832 cites W3139082476 @default.
• W4384694832 cites W3158531670 @default.
• W4384694832 cites W3165499928 @default.
• W4384694832 cites W3175674467 @default.
• W4384694832 cites W3176606713 @default.
• W4384694832 cites W3183264033 @default.
• W4384694832 cites W3206824631 @default.
• W4384694832 cites W4205804651 @default.
• W4384694832 cites W4214767871 @default.
• W4384694832 cites W4220741568 @default.
• W4384694832 cites W4220800505 @default.
• W4384694832 cites W4220882094 @default.
• W4384694832 cites W4220929376 @default.
• W4384694832 cites W4229456824 @default.
• W4384694832 cites W4280539093 @default.
• W4384694832 cites W4288403646 @default.
• W4384694832 cites W4291653105 @default.
• W4384694832 cites W4312191656 @default.
• W4384694832 cites W4312629918 @default.
• W4384694832 cites W4312857285 @default.
• W4384694832 cites W4312913583 @default.
• W4384694832 doi "https://doi.org/10.3389/fnins.2023.1177592" @default.
• W4384694832 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37534034" @default.
• W4384694832 hasPublicationYear "2023" @default.
• W4384694832 type Work @default.
• W4384694832 citedByCount "0" @default.
• W4384694832 crossrefType "journal-article" @default.
• W4384694832 hasAuthorship W4384694832A5028809607 @default.
• W4384694832 hasAuthorship W4384694832A5071014444 @default.
• W4384694832 hasBestOaLocation W43846948321 @default.
• W4384694832 hasConcept C108583219 @default.
• W4384694832 hasConcept C11731999 @default.
• W4384694832 hasConcept C118524514 @default.
• W4384694832 hasConcept C151927369 @default.
• W4384694832 hasConcept C154945302 @default.
• W4384694832 hasConcept C41008148 @default.
• W4384694832 hasConcept C50644808 @default.
• W4384694832 hasConcept C81363708 @default.
• W4384694832 hasConcept C9390403 @default.
• W4384694832 hasConceptScore W4384694832C108583219 @default.
• W4384694832 hasConceptScore W4384694832C11731999 @default.
• W4384694832 hasConceptScore W4384694832C118524514 @default.
• W4384694832 hasConceptScore W4384694832C151927369 @default.
• W4384694832 hasConceptScore W4384694832C154945302 @default.
• W4384694832 hasConceptScore W4384694832C41008148 @default.
• W4384694832 hasConceptScore W4384694832C50644808 @default.
• W4384694832 hasConceptScore W4384694832C81363708 @default.

• next