Utilize este identificador para referenciar este registo:
http://hdl.handle.net/10071/29488
Registo completo
Campo DC | Valor | Idioma |
---|---|---|
dc.contributor.author | Miranda, I. D. S. | - |
dc.contributor.author | Arora, A. | - |
dc.contributor.author | Susskind, Z. | - |
dc.contributor.author | Souza, J. S. A. | - |
dc.contributor.author | Jadhao, M. P. | - |
dc.contributor.author | Villon, L. A. Q. | - |
dc.contributor.author | Dutra, D. L. C. | - |
dc.contributor.author | Lima, P. M. V. | - |
dc.contributor.author | França, F. M. G. | - |
dc.contributor.author | Breternitz Jr., M. | - |
dc.contributor.author | John, L. K. | - |
dc.contributor.editor | Cardoso, J. M. P., Jimborean, A., and Mentens, N. | - |
dc.date.accessioned | 2023-10-30T12:15:11Z | - |
dc.date.issued | 2023 | - |
dc.identifier.citation | Miranda, I. D. S., Arora, A., Susskind, Z., Souza, J. S. A., Jadhao, M. P., Villon, L. A. Q., Dutra, D. L. C., Lima, P. M. V., França, F. M. G., Breternitz Jr., M., & John, L. K. (2023). COIN: Combinational Intelligent Networks. In J. M. P. Cardoso, A. Jimborean, & N. Mentens (Eds.), 2023 IEEE 34th International Conference on Application-specific Systems, Architectures and Processors (ASAP). IEEE. https://doi.org/10.1109/ASAP57973.2023.00016 | - |
dc.identifier.isbn | 979-8-3503-4685-5 | - |
dc.identifier.issn | 2160-0511 | - |
dc.identifier.uri | http://hdl.handle.net/10071/29488 | - |
dc.description.abstract | We introduce Combinational Intelligent Networks (COIN), a machine learning technique that targets edge inference using low-resourced FPGAs or ASICs. COIN is an improvement on LogicWiSARD, a recent weightless neural network that achieves low power, small area, and high throughput. We convert the LogicWiSARD model into a binary neural network, train it using backpropagation, and then convert it to a COIN model. As a result, COIN can achieve higher accuracy than LogicWiSARD or it can require significantly fewer hardware resources when comparing models with similar accuracies. In comparison to a BNN implementation, FINN, small and large COIN models are more energy efficient demonstrating up to 11.5x higher inferences/Joule at similar accuracy. Our tool executes the complete flow, from training to RTL. and is publicly available. | eng |
dc.language.iso | eng | - |
dc.publisher | IEEE | - |
dc.relation | info:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDB%2F50008%2F2020/PT | - |
dc.relation | UIDP/4466/2020 | - |
dc.relation | info:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDB%2F04466%2F2020/PT | - |
dc.relation.ispartof | 2023 IEEE 34th International Conference on Application-specific Systems, Architectures and Processors (ASAP) | - |
dc.rights | embargoedAccess | - |
dc.subject | Weightless neural networks | eng |
dc.subject | LogicWiSARD | eng |
dc.subject | Binary neural networks | eng |
dc.subject | FPGA | eng |
dc.subject | ASIC | eng |
dc.title | COIN: Combinational Intelligent Networks | eng |
dc.type | conferenceObject | - |
dc.event.title | 34th International Conference on Application-specific Systems, Architectures and Processors (ASAP) | - |
dc.event.type | Conferência | pt |
dc.event.location | Porto, Portugal | eng |
dc.event.date | 2023 | - |
dc.peerreviewed | yes | - |
dc.date.updated | 2023-10-30T12:12:16Z | - |
dc.description.version | info:eu-repo/semantics/acceptedVersion | - |
dc.identifier.doi | 10.1109/ASAP57973.2023.00016 | - |
dc.subject.fos | Domínio/Área Científica::Ciências Naturais::Ciências da Computação e da Informação | por |
dc.subject.fos | Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática | por |
dc.date.embargo | 2025-10-01 | - |
iscte.identifier.ciencia | https://ciencia.iscte-iul.pt/id/ci-pub-97244 | - |
Aparece nas coleções: | ISTAR-CRI - Comunicações a conferências internacionais |
Ficheiros deste registo:
Ficheiro | Tamanho | Formato | |
---|---|---|---|
conferenceobject_97244.pdf Restricted Access | 198,71 kB | Adobe PDF | Ver/Abrir Request a copy |
Todos os registos no repositório estão protegidos por leis de copyright, com todos os direitos reservados.