blank Quick help
blank Maintenance news

Scheduled maintenance

Regular maintenance outages:
between 05.00 and 05.15 hrs CET (Monday to Sunday).

Other outages
Availability

2022.02.11

More...
blank News flashes

News Flashes

New version of the European Patent Register – SPC proceedings information in the Unitary Patent Register.

2024-07-24

More...
blank Related links

Extract from the Register of European Patents

EP About this file: EP4196919

EP4196919 - METHOD AND SYSTEM FOR QUANTIZING A NEURAL NETWORK [Right-click to bookmark this link]
StatusRequest for examination was made
Status updated on  19.05.2023
Database last updated on 19.10.2024
FormerThe international publication has been made
Status updated on  21.05.2022
Formerunknown
Status updated on  18.02.2021
Most recent event   Tooltip19.01.2024Change: Validation statespublished on 21.02.2024  [2024/08]
19.01.2024Change - extension statespublished on 21.02.2024  [2024/08]
Applicant(s)For all designated states
Huawei Technologies Co., Ltd.
Huawei Administration Building
Bantian
Longgang District
Shenzhen, Guangdong 518129 / CN
[2023/25]
Inventor(s)01 / SOLODSKIKH, Kirill Igorevich
Shenzhen Guangdong, 518129 / CN
02 / TELEGINA, Anna Dmitrievna
Shenzhen Guangdong, 518129 / CN
03 / CHIKIN, Vladimir Maximovich
Shenzhen Guangdong, 518129 / CN
 [2023/25]
Representative(s)Epping - Hermann - Fischer
Patentanwaltsgesellschaft mbH
Schloßschmidstraße 5
80639 München / DE
[2023/25]
Application number, filing date20851354.913.11.2020
[2023/25]
WO2020RU00601
Filing languageEN
Procedural languageEN
PublicationType: A1 Application with search report
No.:WO2022103291
Date:19.05.2022
Language:EN
[2022/20]
Type: A1 Application with search report 
No.:EP4196919
Date:21.06.2023
Language:EN
The application published by WIPO in one of the EPO official languages on 19.05.2022 takes the place of the publication of the European patent application.
[2023/25]
Search report(s)International search report - published on:EP19.05.2022
ClassificationIPC:G06N3/08
[2023/25]
CPC:
H03M7/3059 (EP); G06N3/084 (EP); H03M7/3082 (EP);
H03M7/70 (EP)
Designated contracting statesAL,   AT,   BE,   BG,   CH,   CY,   CZ,   DE,   DK,   EE,   ES,   FI,   FR,   GB,   GR,   HR,   HU,   IE,   IS,   IT,   LI,   LT,   LU,   LV,   MC,   MK,   MT,   NL,   NO,   PL,   PT,   RO,   RS,   SE,   SI,   SK,   SM,   TR [2023/25]
TitleGerman:VERFAHREN UND SYSTEM ZUR QUANTISIERUNG EINES NEURONALEN NETZES[2023/25]
English:METHOD AND SYSTEM FOR QUANTIZING A NEURAL NETWORK[2023/25]
French:PROCÉDÉ ET SYSTÈME PERMETTANT DE QUANTIFIER UN RÉSEAU NEURONAL[2023/25]
Entry into regional phase14.03.2023National basic fee paid 
14.03.2023Designation fee(s) paid 
14.03.2023Examination fee paid 
Examination procedure14.03.2023Examination requested  [2023/25]
14.03.2023Date on which the examining division has become responsible
09.08.2023Amendment by applicant (claims and/or description)
Fees paidRenewal fee
14.03.2023Renewal fee patent year 03
22.11.2023Renewal fee patent year 04
Opt-out from the exclusive  Tooltip
competence of the Unified
Patent Court
See the Register of the Unified Patent Court for opt-out data
Responsibility for the accuracy, completeness or quality of the data displayed under the link provided lies entirely with the Unified Patent Court.
Cited inInternational search[I]  - AHMED T ELTHAKEB ET AL, "WaveQ: Gradient-Based Deep Quantization of Neural Networks through Sinusoidal Adaptive Regularization", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, (20200229), XP081651983 [I] 1-20 * abstract * * sections 1, 2, 4, 6, 7 *
 [A]  - AHMED T ELTHAKEB ET AL, "SinReQ: Generalized Sinusoidal Regularization for Low-Bitwidth Deep Quantized Training", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, (20190504), XP081542586 [A] 1-20 * abstract * * sections 1, 3-5 *
 [A]  - MAXIM NAUMOV ET AL, "On Periodic Functions as Regularizers for Quantization of Neural Networks", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, (20181124), XP081040812 [A] 1-20 * abstract * * sections 1, 2 *
 [A]  - STEFAN UHLICH ET AL, "Mixed Precision DNNs: All you need is a good parametrization", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, (20190527), XP081663581 [A] 1-20 * abstract * * section 3 *
 [A]  - STEVEN K ESSER ET AL, "Learned Step Size Quantization", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, (20200507), XP081663151 [A] 1-20 * abstract * * section 3.3 *
The EPO accepts no responsibility for the accuracy of data originating from other authorities; in particular, it does not guarantee that it is complete, up to date or fit for specific purposes.