Title:
BRAIN-LIKE NEURAL NETWORK HAVING MEMORY AND INFORMATION ABSTRACTION FUNCTIONS
Document Type and Number:
WIPO Patent Application WO/2021/233180
Kind Code:
A1
Abstract:
A brain-like neural network having memory and information abstraction functions. The brain-like neural network uses operating principles of a hippocampal region of a biological brain and multiple peripheral brain regions for references, and comprises a memory module capable of forming an episodic memory, so that an intelligent body can efficiently recognize objects, perform spatial navigation, reasoning and autonomous decision making, can quickly memorize the characteristics of the objects and perform abstraction and meta-learning, has a strong generalization capability, and is capable of lifelong learning; a synaptic plasticity process is used to adjust a weight, partial differential operation is avoided, computational overhead is lower than that of conventional deep learning, and a basis is provided for the design and application of a neural mimetic chip.
More Like This:
Inventors:
REN HUALONG (CN)
Application Number:
PCT/CN2021/093355
Publication Date:
November 25, 2021
Filing Date:
May 12, 2021
Export Citation:
Assignee:
NEUROCEAN TECH INC (CN)
International Classes:
G06N3/063
Foreign References:
| CN110427536A | 2019-11-08 | |||
| CN103116279A | 2013-05-22 | |||
| CN104809498A | 2015-07-29 | |||
| CN107563505A | 2018-01-09 | |||
| US20190366538A1 | 2019-12-05 |
Attorney, Agent or Firm:
SHENZHEN ZHONGYI UNION INTELLECTUAL PROPERTY AGENCY CO.,LTD. (CN)
Download PDF:
Previous Patent: BRAIN-LIKE VISUAL NEURAL NETWORK HAVING FORWARD LEARNING AND META-LEARNING FUNCTIONS
Next Patent: EMBEDDED AIR CONDITIONER
Next Patent: EMBEDDED AIR CONDITIONER