NEURAL NETWORK LEARNING DEVICE, METHOD, AND PROGRAM

Title:

NEURAL NETWORK LEARNING DEVICE, METHOD, AND PROGRAM

Document Type and Number:

WIPO Patent Application WO/2018/167900

Kind Code:

A1

Abstract:

A large amount of learning data is typically required to perform deep network leaning, making it difficult to achieve learning using a few pieces of data. In order to solve this problem, the neural network learning device according to the present invention is provided with: a feature extraction unit which extracts features from learning data using a learning neural network; an adversarial feature generation unit which generates an adversarial feature from the extracted features using the learning neural network; a pattern recognition unit which calculates a neural network recognition result using the learning data and the adversarial feature; and a network learning unit which performs neural network learning so that the recognition result approaches a desired output.

More Like This:

WO/2016/187500	METHOD AND APPARATUS FOR TRAINING MEMRISTIVE LEARNING SYSTEMS
WO/2024/051183	BACKDOOR DETECTION METHOD BASED ON DECISION SHORTCUT SEARCH
WO/2023/196435	AUTONOMOUS, WORLD-BUILDING, LIFELONG LEARNING AGENTS AND COMPUTING ENGINES FOR GENERAL-PURPOSE INTELLIGENCE

Inventors:

ISHII MASATO (JP)

Application Number:

PCT/JP2017/010560

Publication Date:

September 20, 2018

Filing Date:

March 16, 2017

Export Citation:

Click for automatic bibliography generation Help

Assignee:

NEC CORP (JP)

International Classes:

G06N3/08; G06V10/764; G06V10/774; G06V10/776

Foreign References:

JP2014178229A	2014-09-25
JP2009032273A	2009-02-12
JP2016099507A	2016-05-30