View in EDS

Overview of the Neural Network Compression and Representation (NNR) Standard.

Saved in:
Bibliographic Details
Title: Overview of the Neural Network Compression and Representation (NNR) Standard.
Authors: Kirchhoffer, Heiner1 heiner.kirchhoffer@hhi.fraunhofer.de, Haase, Paul1, Samek, Wojciech1, Muller, Karsten1 karsten.mueller@hhi.fraunhofer.de, Rezazadegan-Tavakoli, Hamed2, Cricri, Francesco2, Aksu, Emre B.2, Hannuksela, Miska M.2, Jiang, Wei3, Wang, Wei3, Liu, Shan3, Jain, Swayambhoo4, Hamidi-Rad, Shahab4, Racape, Fabien4, Bailer, Werner5
Source: IEEE Transactions on Circuits & Systems for Video Technology. May2022, Vol. 32 Issue 5, p3203-3216. 14p.
Subject Terms: *ARTIFICIAL neural networks, LINEAR network coding, NEURAL codes, BINARY codes, BIOLOGICAL neural networks
Abstract: Neural Network Coding and Representation (NNR) is the first international standard for efficient compression of neural networks (NNs). The standard is designed as a toolbox of compression methods, which can be used to create coding pipelines. It can be either used as an independent coding framework (with its own bitstream format) or together with external neural network formats and frameworks. For providing the highest degree of flexibility, the network compression methods operate per parameter tensor in order to always ensure proper decoding, even if no structure information is provided. The NNR standard contains compression-efficient quantization and deep context-adaptive binary arithmetic coding (DeepCABAC) as core encoding and decoding technologies, as well as neural network parameter pre-processing methods like sparsification, pruning, low-rank decomposition, unification, local scaling and batch norm folding. NNR achieves a compression efficiency of more than 97% for transparent coding cases, i.e. without degrading classification quality, such as top-1 or top-5 accuracies. This paper provides an overview of the technical features and characteristics of NNR. [ABSTRACT FROM AUTHOR]
Copyright of IEEE Transactions on Circuits & Systems for Video Technology is the property of IEEE and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Business Source Index
Description
Abstract:Neural Network Coding and Representation (NNR) is the first international standard for efficient compression of neural networks (NNs). The standard is designed as a toolbox of compression methods, which can be used to create coding pipelines. It can be either used as an independent coding framework (with its own bitstream format) or together with external neural network formats and frameworks. For providing the highest degree of flexibility, the network compression methods operate per parameter tensor in order to always ensure proper decoding, even if no structure information is provided. The NNR standard contains compression-efficient quantization and deep context-adaptive binary arithmetic coding (DeepCABAC) as core encoding and decoding technologies, as well as neural network parameter pre-processing methods like sparsification, pruning, low-rank decomposition, unification, local scaling and batch norm folding. NNR achieves a compression efficiency of more than 97% for transparent coding cases, i.e. without degrading classification quality, such as top-1 or top-5 accuracies. This paper provides an overview of the technical features and characteristics of NNR. [ABSTRACT FROM AUTHOR]
ISSN:10518215
DOI:10.1109/TCSVT.2021.3095970