http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png

Energy-Efficient Neural Network Inference through Golomb-Rice Compression of Activations for Edge Devices

Vaddeboina, Mounika (Corresponding author) ; Yilmazer, Alper (Corresponding author)^RWTH* ; Ecker, Wolfgang (Corresponding author)

In
2025 28th International Symposium on Design and Diagnostics of Electronic Circuits & Systems - DDECS : May 5-7 2025 : proceedings / editors: Alberto Bosio (École centrale de Lyon, Institute of Nanotechnology, France), Paolo Bernardi (Politecnico di Torino, Italy), Marcello Traiola (INRIA, France), Vojtech Mrazek (Brno University of Technology, Czechia) ; publisher: IEEE, Seiten/Artikel-Nr: 1-6

2025

Konferenz/Event:IEEE 28. International Symposium on Design and Diagnostics of Electronic Circuits and Systems , Lyon , France , DDECS 2025 , 2025-05-05 - 2025-05-07

Impressum[Piscataway, NJ] : IEEE

Umfang1-6

ISBN979-8-3315-2801-0, 979-8-3315-2802-7, 9798331528010, 9798331528027

Online
DOI: 10.1109/DDECS63720.2025.11006805
10.1109/DDECS63720.2025.11006805

Einrichtungen

1. Lehrstuhl für Software für Systeme auf Silizium (611910)

Dokumenttyp
Contribution to a book/Contribution to a conference proceedings

Format
online, print

Sprache
English

Anmerkung
Peer review status of article unknown

Externe Identnummern
SCOPUS: SCOPUS:2-s2.0-105007545993
WOS Core Collection: WOS:001506891000001

Interne Identnummern
RWTH-2025-06625
Datensatz-ID: 1015896

Beteiligte Länder
Germany