Sub-byte quantization of Mobile Face Recognition Convolutional Neural Networks
| dc.contributor.author | Sebastian Bunda, Luuk Spreeuwers and Chris Zeinstra | |
| dc.contributor.editor | Brömme, Arslan | |
| dc.contributor.editor | Damer, Naser | |
| dc.contributor.editor | Gomez-Barrero, Marta | |
| dc.contributor.editor | Raja, Kiran | |
| dc.contributor.editor | Rathgeb, Christian | |
| dc.contributor.editor | Sequeira Ana F. | |
| dc.contributor.editor | Todisco, Massimiliano | |
| dc.contributor.editor | Uhl, Andreas | |
| dc.date.accessioned | 2022-10-27T10:19:29Z | |
| dc.date.available | 2022-10-27T10:19:29Z | |
| dc.date.issued | 2022 | |
| dc.description.abstract | Converting convolutional neural networks such as MobileNets to a full integer representation is already quite a popular method to reduce the size and computational footprint of classification networks but its effect on face recognition networks is relatively unexplored. This work presents a method to reduce the size of MobileFaceNet using sub-byte quantization of the weights and activations. It was found that 8-bit and 4-bit versions of MobileFaceNet can be obtained with 98.68% and 98.63% accuracy on the LFW dataset which reduces the footprint to 25% and 12.5% of the original weights respectively. Using mixed-precision, an accuracy of 98.17% can be achieved whilst requiring only 10% of the original weight footprint. It is expected that with a larger training dataset, higher accuracies can be achieved. | en |
| dc.identifier.doi | 10.1109/BIOSIG55365.2022.9897025 | |
| dc.identifier.isbn | 978-3-88579-723-4 | |
| dc.identifier.pissn | 1617-5490 | |
| dc.identifier.uri | https://dl.gi.de/handle/20.500.12116/39700 | |
| dc.language.iso | en | |
| dc.publisher | Gesellschaft für Informatik e.V. | |
| dc.relation.ispartof | BIOSIG 2022 | |
| dc.relation.ispartofseries | Lecture Notes in Informatics (LNI) - Proceedings, Volume P-329 | |
| dc.subject | Resource Limited Face Recognition | |
| dc.subject | Deep Neural Networks | |
| dc.subject | QKeras | |
| dc.subject | Sub-byte Quantization | |
| dc.title | Sub-byte quantization of Mobile Face Recognition Convolutional Neural Networks | en |
| dc.type | Text/Conference Paper | |
| gi.citation.endPage | 236 | |
| gi.citation.publisherPlace | Bonn | |
| gi.citation.startPage | 229 | |
| gi.conference.date | 14.-16. September 2022 | |
| gi.conference.location | Darmstadt | |
| gi.conference.sessiontitle | Further Conference Contributions |
Dateien
Originalbündel
1 - 1 von 1
Lade...
- Name:
- 23-BIOSIG_2022_paper_4.pdf
- Größe:
- 1.48 MB
- Format:
- Adobe Portable Document Format