Approximate-computing architectures for motion estimation in HEVC

Alberto Paltrinieri; Riccardo Peloso; Guido Masera; Muhammad Shafique; Maurizio Martina

doi:10.1109/NGCAS.2018.8572080

Approximate-computing architectures for motion estimation in HEVC

Alberto Paltrinieri, Riccardo Peloso, Guido Masera, Muhammad Shafique, Maurizio Martina

Electrical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Video compression deeply relies on motion estimation but this crucial step requires power-hungry computational resources. Starting from an existing architecture capable of calculating a high number of Sum of Absolute Differences (SADs), three possible locations for approximate adders are chosen. At each location, different types of approximate adders are implemented and the results are analyzed in terms of error and power saving.

Original language	English (US)
Title of host publication	2018 New Generation of CAS, NGCAS 2018
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	190-193
Number of pages	4
ISBN (Electronic)	9781538676813
DOIs	https://doi.org/10.1109/NGCAS.2018.8572080
State	Published - Dec 10 2018
Event	2018 New Generation of CAS, NGCAS 2018 - Valletta, Malta Duration: Nov 20 2018 → Nov 23 2018

Publication series

Name	2018 New Generation of CAS, NGCAS 2018

Conference

Conference	2018 New Generation of CAS, NGCAS 2018
Country/Territory	Malta
City	Valletta
Period	11/20/18 → 11/23/18

ASJC Scopus subject areas

Hardware and Architecture
Electrical and Electronic Engineering
Energy Engineering and Power Technology
Instrumentation

Access to Document

10.1109/NGCAS.2018.8572080

Cite this

Approximate-computing architectures for motion estimation in HEVC. / Paltrinieri, Alberto; Peloso, Riccardo; Masera, Guido et al.
2018 New Generation of CAS, NGCAS 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 190-193 8572080 (2018 New Generation of CAS, NGCAS 2018).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Paltrinieri, A, Peloso, R, Masera, G, Shafique, M & Martina, M 2018, Approximate-computing architectures for motion estimation in HEVC. in 2018 New Generation of CAS, NGCAS 2018., 8572080, 2018 New Generation of CAS, NGCAS 2018, Institute of Electrical and Electronics Engineers Inc., pp. 190-193, 2018 New Generation of CAS, NGCAS 2018, Valletta, Malta, 11/20/18. https://doi.org/10.1109/NGCAS.2018.8572080

@inproceedings{29dcd0720d0748dba531dacfb021678b,

title = "Approximate-computing architectures for motion estimation in HEVC",

abstract = "Video compression deeply relies on motion estimation but this crucial step requires power-hungry computational resources. Starting from an existing architecture capable of calculating a high number of Sum of Absolute Differences (SADs), three possible locations for approximate adders are chosen. At each location, different types of approximate adders are implemented and the results are analyzed in terms of error and power saving.",

author = "Alberto Paltrinieri and Riccardo Peloso and Guido Masera and Muhammad Shafique and Maurizio Martina",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE. Copyright: Copyright 2019 Elsevier B.V., All rights reserved.; 2018 New Generation of CAS, NGCAS 2018 ; Conference date: 20-11-2018 Through 23-11-2018",

year = "2018",

month = dec,

day = "10",

doi = "10.1109/NGCAS.2018.8572080",

language = "English (US)",

series = "2018 New Generation of CAS, NGCAS 2018",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "190--193",

booktitle = "2018 New Generation of CAS, NGCAS 2018",

}

TY - GEN

T1 - Approximate-computing architectures for motion estimation in HEVC

AU - Paltrinieri, Alberto

AU - Peloso, Riccardo

AU - Masera, Guido

AU - Shafique, Muhammad

AU - Martina, Maurizio

PY - 2018/12/10

Y1 - 2018/12/10

N2 - Video compression deeply relies on motion estimation but this crucial step requires power-hungry computational resources. Starting from an existing architecture capable of calculating a high number of Sum of Absolute Differences (SADs), three possible locations for approximate adders are chosen. At each location, different types of approximate adders are implemented and the results are analyzed in terms of error and power saving.

AB - Video compression deeply relies on motion estimation but this crucial step requires power-hungry computational resources. Starting from an existing architecture capable of calculating a high number of Sum of Absolute Differences (SADs), three possible locations for approximate adders are chosen. At each location, different types of approximate adders are implemented and the results are analyzed in terms of error and power saving.

UR - http://www.scopus.com/inward/record.url?scp=85060254481&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85060254481&partnerID=8YFLogxK

U2 - 10.1109/NGCAS.2018.8572080

DO - 10.1109/NGCAS.2018.8572080

M3 - Conference contribution

AN - SCOPUS:85060254481

T3 - 2018 New Generation of CAS, NGCAS 2018

SP - 190

EP - 193

BT - 2018 New Generation of CAS, NGCAS 2018

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2018 New Generation of CAS, NGCAS 2018

Y2 - 20 November 2018 through 23 November 2018

ER -

Approximate-computing architectures for motion estimation in HEVC

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this