Adaptive Input Normalization for Quantized Neural Networks

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21240%2F24%3A00375936" target="_blank" >RIV/68407700:21240/24:00375936 - isvavai.cz</a>
Result on the web
<a href="https://doi.org/10.1109/DDECS60919.2024.10508927" target="_blank" >https://doi.org/10.1109/DDECS60919.2024.10508927</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1109/DDECS60919.2024.10508927" target="_blank" >10.1109/DDECS60919.2024.10508927</a>

Result language
angličtina
Original language name
Adaptive Input Normalization for Quantized Neural Networks
Original language description
Neural networks with quantized activation functions cannot adapt the quantization at the input of their first layer. Preprocessing is therefore required to adapt the range of input data to the quantization range. Such preprocessing usually includes an activation-wise linear transformation and is steered by the properties of the training set. We suggest to include the linear transform into the training process. We document that it improves accuracy, requires the same resources as standard preprocessing, plays a role in network pruning, and is reasonably stable with respect to initialization.
Czech name
—
Czech description
—

Project
—
Continuities
I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year
2024
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection
Proceedings of the 27th International Symposium on Design and Diagnostics of Electronic Circuits & Systems
ISBN
979-8-3503-5934-3
ISSN
—
e-ISSN
—
Number of pages
6
Pages from-to
130-135
Publisher name
IEEE
Place of publication
Piscataway
Event location
Kielce
Event date
Apr 3, 2024
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
001227439800025

Similar results(10)