Repository logo
 
Publication

A full featured configurable accelerator for object detection with YOLO

2021-05-19Journal article Restricted access
Simple item page
dc.contributor.authorPestana, Daniel
dc.contributor.authorMiranda, Pedro R.
dc.contributor.authorLopes, João D.
dc.contributor.authorDuarte, Rui
dc.contributor.authorVéstias, Mário
dc.contributor.authorNeto, Horácio C
dc.contributor.authorDe Sousa, Jose
dc.date.accessioned2021-09-07T09:41:09Z
dc.date.available2021-09-07T09:41:09Z
dc.date.issued2021-05-19
dc.description.abstractObject detection and classification is an essential task of computer vision. A very efficient algorithm for detection and classification is YOLO (You Look Only Once). We consider hardware architectures to run YOLO in real-time on embedded platforms. Designing a new dedicated accelerator for each new version of YOLO is not feasible given the fast delivery of new versions. This work's primary goal is to design a configurable and scalable core for creating specific object detection and classification systems based on YOLO, targeting embedded platforms. The core accelerates the execution of all the algorithm steps, including pre-processing, model inference and post-processing. It considers a fixed-point format, linearised activation functions, batch-normalisation, folding, and a hardware structure that exploits most of the available parallelism in CNN processing. The proposed core is configured for real-time execution of YOLOv3-Tiny and YOLOv4-Tiny, integrated into a RISC-V-based system-on-chip architecture and prototyped in an UltraScale XCKU040 FPGA (Field Programmable Gate Array). The solution achieves a performance of 32 and 31 frames per second for YOLOv3-Tiny and YOLOv4-Tiny, respectively, with a 16-bit fixed-point format. Compared to previous proposals, it improves the frame rate at a higher performance efficiency. The performance, area efficiency and configurability of the proposed core enable the fast development of real-time YOLO-based object detectors on embedded systems.pt_PT
dc.description.versioninfo:eu-repo/semantics/publishedVersionpt_PT
dc.identifier.citationPESTANA, Daniel; [et al] – A full featured configurable accelerator for object detection with YOLO. IEEE Access. ISSN 2169-3536. Vol. 9 (2021), pp. 75864-75877pt_PT
dc.identifier.doi10.1109/ACCESS.2021.3081818pt_PT
dc.identifier.issn2169-3536
dc.identifier.urihttp://hdl.handle.net/10400.21/13689
dc.language.isoengpt_PT
dc.peerreviewedyespt_PT
dc.publisherIEEEpt_PT
dc.relation.publisherversionhttps://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9435338&tag=1pt_PT
dc.subjectObject detectionpt_PT
dc.subjectConvolutional neural networkpt_PT
dc.subjectFPGApt_PT
dc.subjectLightweight YOLOpt_PT
dc.titleA full featured configurable accelerator for object detection with YOLOpt_PT
dc.typejournal article
dspace.entity.typePublication
oaire.citation.endPage75877pt_PT
oaire.citation.startPage75864pt_PT
oaire.citation.titleIEEE Accesspt_PT
oaire.citation.volume9pt_PT
person.familyNameDuarte
person.familyNameVéstias
person.familyNameCláudio de Campos Neto
person.familyNamede Sousa
person.givenNameRui
person.givenNameMário
person.givenNameHorácio
person.givenNameJose
person.identifier.ciencia-idB91E-770F-19A3
person.identifier.ciencia-id4717-C2C7-3F2C
person.identifier.ciencia-id9915-3BDF-5C35
person.identifier.ciencia-idBE18-E262-E0EC
person.identifier.orcid0000-0002-7060-4745
person.identifier.orcid0000-0001-8556-4507
person.identifier.orcid0000-0002-3621-8322
person.identifier.orcid0000-0001-7525-7546
person.identifier.ridI-4402-2015
person.identifier.ridH-9953-2012
person.identifier.ridL-6859-2015
person.identifier.scopus-author-id24823991600
person.identifier.scopus-author-id14525867300
person.identifier.scopus-author-id7102813024
rcaap.rightsclosedAccesspt_PT
rcaap.typearticlept_PT
relation.isAuthorOfPublicationf2b4b9e6-6c89-48c7-bc83-62d2e98a787b
relation.isAuthorOfPublicationa7d22b29-c961-45ac-bc09-cd5e1002f1e8
relation.isAuthorOfPublication38334d5e-83e8-494c-a9e0-396299376d97
relation.isAuthorOfPublicationd98a4d45-2d45-42ec-9f1d-14775723709b
relation.isAuthorOfPublication.latestForDiscoverya7d22b29-c961-45ac-bc09-cd5e1002f1e8

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
A full_MVestias.pdf
Size:
1.27 MB
Format:
Adobe Portable Document Format
Download

Collections

ISEL - Eng. Elect. Tel. Comp. - Artigos