Search: id:"swepub:oai:lup.lub.lu.se:64c5c8d0-749c-430e-9ecd-02c3b48b2039" >
Compression of Acti...
Compression of Activation Signals from Split Deep Neural Network
-
- Brito, Flavio (author)
- Lund University,Lunds universitet,Bredbandskommunikation,Forskargrupper vid Lunds universitet,Broadband Communication,Lund University Research Groups,Ericsson Research
-
- Silva, Lucas (author)
- Federal University of Pará
-
- Ramalho, Leonardo (author)
- Federal University of Pará
-
show more...
-
- Lins, Silvia (author)
- Ericsson Research
-
- Linder, Neiva (author)
- Ericsson Research
-
- Klautau, Aldebaro (author)
- Federal University of Pará
-
Moraes, Igor M. (editor)
-
Campista, Miguel Elias M. (editor)
-
Ghamri-Doudane, Yacine (editor)
-
Luis Henrique M. K. Costa, Costa (editor)
-
Rubinstein, Marcelo G. (editor)
-
show less...
-
(creator_code:org_t)
- 2022
- 2022
- English.
-
In: 2022 IEEE Latin-American Conference on Communications, LATINCOM 2022. - 9781665482257
- Related links:
-
http://dx.doi.org/10...
-
show more...
-
https://lup.lub.lu.s...
-
https://doi.org/10.1...
-
show less...
Abstract
Subject headings
Close
- The use of artificial neural networks for the purpose of image classification, together with the advancement in computational capabilities of edge devices, plays an important role in the new emerging 5G use case scenarios. However, one of the main challenges of applications involving the use of these networks in edge devices is still the limitation of computational resources. An alternative for saving resources and promoting privacy are the split learning (or split inference) techniques, in which a deep neural network is cut into two parts and executed in distinct devices. Most of these techniques rely on sending the activation signals (output of the cut layer) through the communication channel. This work proposes a new compression algorithm for decreasing the bit rate required for the transmission of the activation signals (or simply 'scores'). The presented results demonstrate that the transmission rate can be decreased without hurting the neural network accuracy.
Subject headings
- NATURVETENSKAP -- Data- och informationsvetenskap -- Datorteknik (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Computer Engineering (hsv//eng)
Keyword
- deep neural networks
- image classification
- scores compression
- Split learning
Publication and Content Type
- kon (subject category)
- ref (subject category)
Find in a library
To the university's database
- By the author/editor
-
Brito, Flavio
-
Silva, Lucas
-
Ramalho, Leonard ...
-
Lins, Silvia
-
Linder, Neiva
-
Klautau, Aldebar ...
-
show more...
-
Moraes, Igor M.
-
Campista, Miguel ...
-
Ghamri-Doudane, ...
-
Luis Henrique M. ...
-
Rubinstein, Marc ...
-
show less...
- About the subject
-
- NATURAL SCIENCES
-
NATURAL SCIENCES
-
and Computer and Inf ...
-
and Computer Enginee ...
- Articles in the publication
-
2022 IEEE Latin- ...
- By the university
-
Lund University