WO2013104938A3 - Training of a convolutional neural net - Google Patents

Training of a convolutional neural net Download PDF

Info

Publication number
WO2013104938A3
WO2013104938A3 PCT/HU2013/000006 HU2013000006W WO2013104938A3 WO 2013104938 A3 WO2013104938 A3 WO 2013104938A3 HU 2013000006 W HU2013000006 W HU 2013000006W WO 2013104938 A3 WO2013104938 A3 WO 2013104938A3
Authority
WO
WIPO (PCT)
Prior art keywords
training
neural network
image
sub
network
Prior art date
Application number
PCT/HU2013/000006
Other languages
French (fr)
Other versions
WO2013104938A2 (en
Inventor
Gábor BAYER
Original Assignee
77 Elektronika Muszeripari Kft.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 77 Elektronika Muszeripari Kft. filed Critical 77 Elektronika Muszeripari Kft.
Publication of WO2013104938A2 publication Critical patent/WO2013104938A2/en
Publication of WO2013104938A3 publication Critical patent/WO2013104938A3/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
    • G06V10/449—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters
    • G06V10/451—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters with interaction between the filter responses, e.g. cortical complex cells
    • G06V10/454—Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]

Abstract

The invention is, on the one hand, method for training a neural network - the neural network being adapted for generating probability maps - each associated with a respective category of elements in an image - on a basis of visual information detectable in the image of a body fluid sample, the probability map comprising presence probability values of an element of the given category, and - the neural network being a convolution neural network and comprising an input layer (20) detecting inputs from the image, at least one intermediate layer (21) and an output layer (23) providing the presence probability values and comprising sub-layers in a number corresponding to the number of the categories, The method is characterised by - carrying out the training by means of training images characterised by an element associated with only up to one category and being smaller than the image, for a sub-network of the neural network, the sub-network corresponding to the size of the training image and comprising neurons (24) of the input layer detecting inputs from the training image, and neurons (24) - linked to these directly or indirectly by weights - of the at least one intermediate layer and the output layer, - and generating the complete neural network with weights (25) resulting from the training of the sub-network. On the other hand the invention is a neural network trained by the above method.
PCT/HU2013/000006 2012-01-11 2013-01-09 Neural network and a method for teaching thereof WO2013104938A2 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
HUP1200018 2012-01-11
HU1200018A HUP1200018A2 (en) 2012-01-11 2012-01-11 Method of training a neural network, as well as a neural network

Publications (2)

Publication Number Publication Date
WO2013104938A2 WO2013104938A2 (en) 2013-07-18
WO2013104938A3 true WO2013104938A3 (en) 2013-11-07

Family

ID=89990573

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/HU2013/000006 WO2013104938A2 (en) 2012-01-11 2013-01-09 Neural network and a method for teaching thereof

Country Status (2)

Country Link
HU (1) HUP1200018A2 (en)
WO (1) WO2013104938A2 (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104751162A (en) * 2015-04-03 2015-07-01 哈尔滨工业大学 Hyperspectral remote sensing data feature extraction method based on convolution neural network
US20170256038A1 (en) * 2015-09-24 2017-09-07 Vuno Korea, Inc. Image Generating Method and Apparatus, and Image Analyzing Method
CN105528638B (en) * 2016-01-22 2018-04-24 沈阳工业大学 The method that gray relative analysis method determines convolutional neural networks hidden layer characteristic pattern number
CN110709749B (en) 2017-06-09 2021-10-26 电子慕泽雷帕里公司 Combined bright field and phase contrast microscope system and image processing apparatus equipped therewith
TWI717655B (en) * 2018-11-09 2021-02-01 財團法人資訊工業策進會 Feature determination apparatus and method adapted to multiple object sizes
CN113383225A (en) * 2018-12-26 2021-09-10 加利福尼亚大学董事会 System and method for propagating two-dimensional fluorescence waves onto a surface using deep learning
US11803963B2 (en) 2019-02-01 2023-10-31 Sartorius Bioanalytical Instruments, Inc. Computational model for analyzing images of a biological specimen
WO2020225580A1 (en) 2019-05-08 2020-11-12 77 Elektronika Műszeripari Kft. Image taking method, image analysis method, method for training an image analysis neural network, and image analysis neural network
CN116406468A (en) * 2020-11-17 2023-07-07 赛多利斯生物分析仪器有限公司 Computational model for analyzing images of biological specimens
CN113420813B (en) * 2021-06-23 2023-11-28 北京市机械工业局技术开发研究所 Diagnostic method for particulate matter filter cotton state of vehicle tail gas detection equipment
CN117520753B (en) * 2024-01-05 2024-04-05 河北中体善建体育产业有限公司 Early warning system and method for ice and snow sports

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5052043A (en) 1990-05-07 1991-09-24 Eastman Kodak Company Neural network with back propagation controlled through an output confidence measure
US5796410A (en) 1990-06-12 1998-08-18 Lucent Technologies Inc. Generation and use of defective images in image analysis
US5903884A (en) 1995-08-08 1999-05-11 Apple Computer, Inc. Method for training a statistical classifier with reduced tendency for overfitting
US6876966B1 (en) 2000-10-16 2005-04-05 Microsoft Corporation Pattern recognition training method and apparatus using inserted noise followed by noise reduction
US7130776B2 (en) 2002-03-25 2006-10-31 Lockheed Martin Corporation Method and computer program product for producing a pattern recognition training set
DE10245834A1 (en) 2002-10-01 2004-04-15 Siemens Ag Method for generating learning and / or test samples
US7418128B2 (en) 2003-07-31 2008-08-26 Microsoft Corporation Elastic distortions for automatic generation of labeled data
CN100377171C (en) 2004-08-13 2008-03-26 富士通株式会社 Method and apparatus for generating deteriorated numeral image
US7333963B2 (en) 2004-10-07 2008-02-19 Bernard Widrow Cognitive memory and auto-associative neural network based search engine for computer and network located images and photographs

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
BARBANO P E ET AL: "Toward Automatic Phenotyping of Developing Embryos From Videos", IEEE TRANSACTIONS ON IMAGE PROCESSING, IEEE SERVICE CENTER, PISCATAWAY, NJ, US, vol. 14, no. 9, 1 September 2005 (2005-09-01), pages 1360 - 1371, XP011137594, ISSN: 1057-7149, DOI: 10.1109/TIP.2005.852470 *
LECUN Y ET AL: "GRADIENT-BASED LEARNING APPLIED TO DOCUMENT RECOGNITION", PROCEEDINGS OF THE IEEE, IEEE. NEW YORK, US, vol. 86, no. 11, 1 November 1998 (1998-11-01), pages 2278 - 2323, XP000875095, ISSN: 0018-9219, DOI: 10.1109/5.726791 *
QING ZHENG ET AL: "Direct neural network application for automated cell recognition", CYTOMETRY, vol. 57A, no. 1, 1 January 2004 (2004-01-01), pages 1 - 9, XP055076102, ISSN: 0196-4763, DOI: 10.1002/cyto.a.10106 *
RÉGIS VAILLANT ET AL: "An Original Approach for the Localization of Objects in Images", 1 January 1993 (1993-01-01), pages 1 - 16, XP055077236, Retrieved from the Internet <URL:http://yann.lecun.com/exdb/publis/pdf/vaillant-monrocq-lecun-94.pdf> [retrieved on 20130830] *
YAN LE CUN ET AL: "Modeles Connexionnistes de l'apprentissage", INTELLECTICA, SPECIAL ISSUE ON "APPRENTISSAGE ET MACHINE", 1 January 1987 (1987-01-01), pages 1 - 16, XP055077293, Retrieved from the Internet <URL:http://yann.lecun.com/exdb/publis/psgz/lecun-87b.ps.gz> [retrieved on 20130902] *

Also Published As

Publication number Publication date
WO2013104938A2 (en) 2013-07-18
HUP1200018A2 (en) 2013-07-29

Similar Documents

Publication Publication Date Title
WO2013104938A3 (en) Training of a convolutional neural net
Andréassian et al. HESS Opinions" Crash tests for a standardized evaluation of hydrological models"
Cardillo The life‐history basis of latitudinal diversity gradients: how do species traits vary from the poles to the equator?
Banyard et al. Peste des petits ruminants virus, eastern Asia
MX2009004550A (en) Separation filter deciding device and tire inspecting device.
EP2413259A3 (en) Methods and systems for test automation of forms in web applications
EP2453372A3 (en) Information processing apparatus and method, information processing system, and program
CN108168577A (en) MEMS gyro random error compensation method based on BP neural network
Barber et al. 30 Pluralism explains diversity in the Coral Triangle
CN107016440B (en) The multiresolution deep neural network intelligent diagnosing method of machine driving failure
WO2009038430A3 (en) Character inputting device
Ilyukhin et al. Modernized algorithm of neural network initial weighting factors during the diagnosis of diesel engine faults
Hai et al. Changes in community composition of tropical evergreen forests during succession in ta dung national park, central highlands of vietnam
McRoberts et al. The evolution of a process for selecting and prioritizing plant diseases for recovery plans
Basu A study of the dynamics and genetics of COVID-19 through machine learning
Monti Poisoning attacks against banking fraud detection systems
Shen et al. Glitch classification and clustering for ligo with deep transfer learning
Gharib Shirangi et al. Selection of Representative Models for Decision Making and Optimization under Geological Uncertainty
Oosterbeek et al. The impact of cash transfers on school enrollment: Evidence from Ecuador.
Baker et al. P1-388 Transmission of pandemic influenza a (H1N1) on a passenger aircraft
Shakil et al. Investigation of Air Effluence Using IoT and Machine Learning
Martyn et al. Fever and weight loss in a patient with HIV
Krishna et al. Publication of nomenclatural novelties
Sapelko et al. Paleolimnological approach to the separation of the effects of anthropogenic and natural factors on the lake ecosystems
Aman-Oloniyo et al. P1-386 Investigation of unexplained sporadic deaths in Ajah-Ilaje community: Lagos, July 2008

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13707027

Country of ref document: EP

Kind code of ref document: A2

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)
122 Ep: pct application non-entry in european phase

Ref document number: 13707027

Country of ref document: EP

Kind code of ref document: A2