Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020068498 - DATA COMPRESSION USING INTEGER NEURAL NETWORKS

Publication Number WO/2020/068498
Publication Date 02.04.2020
International Application No. PCT/US2019/051624
International Filing Date 18.09.2019
IPC
G06N 3/04 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
04Architecture, e.g. interconnection topology
G06N 3/08 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
H04N 19/00 2014.01
HELECTRICITY
04ELECTRIC COMMUNICATION TECHNIQUE
NPICTORIAL COMMUNICATION, e.g. TELEVISION
19Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
G06N 3/063 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
06Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
063using electronic means
CPC
G06N 3/0454
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
04Architectures, e.g. interconnection topology
0454using a combination of multiple neural nets
G06N 3/063
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
06Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
063using electronic means
G06N 3/084
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
084Back-propagation
Applicants
  • GOOGLE LLC [US]/[US]
Inventors
  • JOHNSTON, Nicholas
  • BALLE, Johannes
Agents
  • TREILHARD, John
  • PORTNOV, Michael
Priority Data
62/737,85227.09.2018US
Publication Language English (EN)
Filing Language English (EN)
Designated States
Title
(EN) DATA COMPRESSION USING INTEGER NEURAL NETWORKS
(FR) COMPRESSION DE DONNÉES À L'AIDE DE RÉSEAUX NEURONAUX ENTIERS
Abstract
(EN)
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for reliably performing data compression and data decompression across a wide variety of hardware and software platforms by using integer neural networks. In one aspect, there is provided a method for entropy encoding data which defines a sequence comprising a plurality of components, the method comprising: for each component of the plurality of components: processing an input comprising: (i) a respective integer representation of each of one or more components of the data which precede the component in the sequence, (ii) an integer representation of one or more respective latent variables characterizing the data, or (iii) both, using an integer neural network to generate data defining a probability distribution over the predetermined set of possible code symbols for the component of the data.
(FR)
La présente invention concerne des procédés, des systèmes et un appareil, y compris des programmes informatiques codés sur un support d'informations informatique, qui permettent de réaliser de manière fiable une compression de données et une décompression de données sur une grande variété de plateformes matérielles et logicielles en utilisant des réseaux neuronaux entiers. Selon un aspect, l'invention concerne un procédé de codage entropique de données qui définit une séquence comprenant une pluralité de composants, le procédé consistant : pour chaque composant de la pluralité de composants : à traiter une entrée comprenant : (i) une représentation entière respective de chacun d'au moins un composant des données qui précèdent le composant dans la séquence, (ii) une représentation entière d'au moins une variable latente respective caractérisant les données, ou (iii) les deux, à l'aide d'un réseau neuronal entier pour générer des données définissant une distribution de probabilité sur l'ensemble prédéterminé de symboles de code possibles pour le composant des données.
Latest bibliographic data on file with the International Bureau