Show simple item record

dc.contributor.authorCaro Roca, Martí
dc.contributor.authorTabani, Hamid
dc.contributor.authorAbella Ferrer, Jaume
dc.contributor.authorMoll Echeto, Francisco de Borja
dc.contributor.authorMorancho Llena, Enrique
dc.contributor.authorCanal Corretger, Ramon
dc.contributor.authorAltet Sanahujes, Josep
dc.contributor.authorCalomarde Palomino, Antonio
dc.contributor.authorCazorla Almeida, Francisco Javier
dc.contributor.authorRubio Romano, Antonio
dc.contributor.authorFontova Muste, Pau
dc.contributor.authorFornt Mas, Jordi
dc.contributor.otherUniversitat Politècnica de Catalunya. Doctorat en Arquitectura de Computadors
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament d'Enginyeria Electrònica
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament d'Arquitectura de Computadors
dc.contributor.otherUniversitat Politècnica de Catalunya. Doctorat en Enginyeria Electrònica
dc.contributor.otherBarcelona Supercomputing Center
dc.date.accessioned2023-04-20T09:02:44Z
dc.date.issued2023-05
dc.identifier.citationCaro, M. [et al.]. An automotive case study on the limits of approximation for object detection. "Journal of systems architecture", Maig 2023, vol. 138, article 102872, p. 1-14.
dc.identifier.issn1383-7621
dc.identifier.otherhttps://arxiv.org/abs/2304.06327
dc.identifier.urihttp://hdl.handle.net/2117/386458
dc.description.abstractThe accuracy of camera-based object detection (CBOD) built upon deep learning is often evaluated against the real objects in frames only. However, such simplistic evaluation ignores the fact that many unimportant objects are small, distant, or background, and hence, their misdetections have less impact than those for closer, larger, and foreground objects in domains such as autonomous driving. Moreover, sporadic misdetections are irrelevant since confidence on detections is typically averaged across consecutive frames, and detection devices (e.g. cameras, LiDARs) are often redundant, thus providing fault tolerance. This paper exploits such intrinsic fault tolerance of the CBOD process, and assesses in an automotive case study to what extent CBOD can tolerate approximation coming from multiple sources such as lower precision arithmetic, approximate arithmetic units, and even random faults due to, for instance, low voltage operation. We show that the accuracy impact of those sources of approximation is within 1% of the baseline even when considering the three approximate domains simultaneously, and hence, multiple sources of approximation can be exploited to build highly efficient accelerators for CBOD in cars.
dc.description.sponsorshipThis work is partially funded by the DRAC project, which is co-financed by the European Union Regional Development Fund within the framework of the ERDF Operational Program of Catalonia 2014–2020 with a grant of 50% of total cost eligible. This work has also been partially supported by the Spanish Ministry of Science and Innovation under grant PID2019-107255GB.
dc.format.extent14 p.
dc.language.isoeng
dc.rightsAttribution-NonCommercial-NoDerivatives 4.0 International
dc.rights.urihttp://creativecommons.org/licenses/by-nc-nd/4.0/
dc.subjectÀrees temàtiques de la UPC::Informàtica::Arquitectura de computadors
dc.subject.lcshEmbedded computer systems
dc.subject.lcshNeural networks (Computer science)
dc.subject.lcshAutonomous vehicles
dc.subject.otherEnergy efficiency
dc.subject.otherNeural network application
dc.subject.otherAutonomous driving
dc.titleAn automotive case study on the limits of approximation for object detection
dc.typeArticle
dc.subject.lemacOrdinadors immersos, Sistemes d'
dc.subject.lemacXarxes neuronals (Informàtica)
dc.subject.lemacVehicles autònoms
dc.contributor.groupUniversitat Politècnica de Catalunya. EFRICS - Efficient and Robust Integrated Circuits and Systems
dc.contributor.groupUniversitat Politècnica de Catalunya. PM - Programming Models
dc.contributor.groupUniversitat Politècnica de Catalunya. CRAAX - Centre de Recerca d'Arquitectures Avançades de Xarxes
dc.identifier.doi10.1016/j.sysarc.2023.102872
dc.description.peerreviewedPeer Reviewed
dc.relation.publisherversionhttps://www.sciencedirect.com/science/article/pii/S1383762123000516
dc.rights.accessRestricted access - publisher's policy
local.identifier.drac35704041
dc.description.versionPostprint (author's final draft)
dc.relation.projectidinfo:eu-repo/grantAgreement/AEI/Plan Estatal de Investigación Científica y Técnica y de Innovación 2017-2020/PID2019-107255GB-C21/ES/BSC - COMPUTACION DE ALTAS PRESTACIONES
dc.date.lift2025-04-05
local.citation.authorCaro, M.; Tabani, H.; Abella, J.; Moll, F.; Morancho, E.; Canal, R.; Altet, J.; Calomarde, A.; Cazorla, F. J.; Rubio, A.; Fontova, P.; Fornt, J.
local.citation.publicationNameJournal of systems architecture
local.citation.volume138
local.citation.numberarticle 102872
local.citation.startingPage1
local.citation.endingPage14


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record