Mostra el registre d'ítem simple

dc.contributor.authorOro Garcia, David
dc.contributor.authorFernandez Tena, Carles
dc.contributor.authorMartorell Bofill, Xavier
dc.contributor.authorHernando Pericás, Francisco Javier
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament d'Arquitectura de Computadors
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament de Teoria del Senyal i Comunicacions
dc.date.accessioned2016-11-02T14:01:14Z
dc.date.issued2016
dc.identifier.citationOro, D., Fernandez, C., Martorell, X., Hernando, J. Work-efficient parallel non-maximum suppression for embedded GPU architectures. A: IEEE International Conference on Acoustics, Speech, and Signal Processing. "2016 IEEE International Conference on Acoustics, Speech, and Signal Processing: proceedings: March 20-25, 2016: Shanghai International Convention Center: Shanghai, China". Shanghai: Institute of Electrical and Electronics Engineers (IEEE), 2016, p. 1026-1030.
dc.identifier.isbn978-1-4799-9988-0
dc.identifier.urihttp://hdl.handle.net/2117/91351
dc.description.abstractWith the emergence of GPU computing, deep neural networks have become a widely used technique for advancing research in the field of image and speech processing. In the context of object and event detection, slidingwindow classifiers require to choose the best among all positively discriminated candidate windows. In this paper, we introduce the first GPU-based non-maximum suppression (NMS) algorithm for embedded GPU architectures. The obtained results show that the proposed parallel algorithm reduces the NMS latency by a wide margin when compared to CPUs, even clocking the GPU at 50% of its maximum frequency on an NVIDIA Tegra K1. In this paper, we show results for object detection in images. The proposed technique is directly applicable to speech segmentation tasks such as speaker diarization.
dc.format.extent5 p.
dc.language.isoeng
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)
dc.subjectÀrees temàtiques de la UPC::Informàtica::Sistemes d'informació
dc.subjectÀrees temàtiques de la UPC::Informàtica
dc.subject.lcshEmbedded computer systems
dc.subject.lcshInformation display systems
dc.subject.otherEmbedded systems
dc.subject.otherGraphics processing units
dc.subject.otherParallel algorithms
dc.subject.otherWork-efficient parallel nonmaximum suppression
dc.subject.otherEmbedded GPU architectures
dc.subject.otherImage processing
dc.subject.otherSpeech processing
dc.subject.otherDeep neural networks
dc.subject.otherNMS latency
dc.subject.otherPositively discriminated candidate windows
dc.subject.otherParallel algorithm
dc.subject.otherNVIDIA Tegra JC1
dc.subject.otherSpeech segmentation tasks
dc.subject.otherSpeaker diarization
dc.titleWork-efficient parallel non-maximum suppression for embedded GPU architectures
dc.typeConference report
dc.subject.lemacSistemes incrustats (Informàtica)
dc.subject.lemacVisualització (Informàtica)
dc.contributor.groupUniversitat Politècnica de Catalunya. CAP - Grup de Computació d'Altes Prestacions
dc.contributor.groupUniversitat Politècnica de Catalunya. VEU - Grup de Tractament de la Parla
dc.identifier.doi10.1109/ICASSP.2016.7471831
dc.description.peerreviewedPeer Reviewed
dc.relation.publisherversionhttp://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=7471831
dc.rights.accessRestricted access - publisher's policy
local.identifier.drac18765151
dc.description.versionPostprint (published version)
dc.relation.projectidinfo:eu-repo/grantAgreement/EC/H2020/644312/EU/Heterogeneous Secure Multi-level Remote Acceleration Service for Low-Power Integrated Systems and Devices/RAPID
dc.date.lift10000-01-01
local.citation.authorOro, D.; Fernandez, C.; Martorell, X.; Hernando, J.
local.citation.contributorIEEE International Conference on Acoustics, Speech, and Signal Processing
local.citation.pubplaceShanghai
local.citation.publicationName2016 IEEE International Conference on Acoustics, Speech, and Signal Processing: proceedings: March 20-25, 2016: Shanghai International Convention Center: Shanghai, China
local.citation.startingPage1026
local.citation.endingPage1030


Fitxers d'aquest items

Imatge en miniatura

Aquest ítem apareix a les col·leccions següents

Mostra el registre d'ítem simple