Brain-Inspired Neuronal Model of Attention and Memory

受大脑启发的注意力和记忆神经元模型

基本信息

批准号：
EP/D036364/1
负责人：
Roman Borisyuk
金额：
$ 19.77万
依托单位：
Plymouth University
依托单位国家：
英国
项目类别：
Research Grant
财政年份：
2006
资助国家：
英国
起止时间：
2006 至无数据
项目状态：
已结题

来源：
https://gtr.ukri.org/projects?ref=EP%2FD036364%2F1
关键词：
Brain Inspired Neuronal Model Attention

项目摘要

Attention is necessary and vital for living organisms due to the limited processing capability of the visual system which precludes the rapid analysis of the whole visual scene. Selective visual attention is a cognitive process that allows a living organism to extract from the incoming visual information the part that is most important at a given moment and that should be processed in more detail. For example, detailed processing of the extracted information can include novelty detection and allocation of a novel object to memory.In this project a large-scale brain-inspired model of hierarchically organised spiking neurons will be developed, that solves the problem of consecutive selection of objects by combining object oriented attention, memory, and novelty detection. Since we believe that the brain does not invent a special processing mechanism for each cognitive function but adapts similar mechanisms for a particular type of processing, it is a challenge to develop a model based on a small set of general principles of information processing (e.g. synchronisation, adaptation of natural frequencies, resonance amplitude increase). We believe that these theoretical principles are the key to the performance of the biological brain and within the proposed research will be implemented for the first time in combined model of attention and memory. Such developments offer great potential, both in shedding fresh light on the basic mechanisms underpinning information processing in the brain and in the design of a new generation of computational devices, cognitive robots, etc.

由于视觉系统的处理能力有限，因此对活生物体的关注是必要的，至关重要，这阻止了整个视觉场景的快速分析。选择性视觉关注是一个认知过程，它允许活生物体从传入的视觉信息中提取出在给定时刻最重要的部分，应更详细地处理。例如，提取的信息的详细处理可以包括新颖性检测和分配新的对象为记忆。在该项目中，将开发出大规模的脑启发的大脑启发的模型，这些模型将开发出层次有组织的尖峰神经元，该模型通过结合对象的注意力，记忆，记忆和新颖的检测来解决连续选择对象的问题的问题。由于我们认为大脑没有为每个认知功能发明特殊的处理机制，而是针对特定类型的处理适应了相似的机制，因此开发基于信息处理的一系列通用原理的模型是一个挑战（例如，同步，适应性，天然频率的适应性，弥补，共振效果振幅增加）。我们认为，这些理论原理是生物学大脑表现的关键，在拟议的研究中将首次在关注和记忆的结合模型中实施。这些发展具有巨大的潜力，既可以阐明大脑中信息处理的基本机制以及新一代计算设备，认知机器人等的设计。