[1]王莉莉 张少白.一种基于声音分类学习的神经模型研究[J].计算机技术与发展,2011,(12):127-130.
 WANG Li-li,ZHANG Shao-bai.Study of a Neural Model Based on Sound Category Learning[J].,2011,(12):127-130.
点击复制

一种基于声音分类学习的神经模型研究()
分享到:

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2011年12期
页码:
127-130
栏目:
智能、算法、系统工程
出版日期:
1900-01-01

文章信息/Info

Title:
Study of a Neural Model Based on Sound Category Learning
文章编号:
1673-629X(2011)12-0127-04
作者:
王莉莉 张少白
南京邮电大学计算机学院
Author(s):
WANG Li-liZHANG Shao-bai
Computer College, Nanjing University of Posts & Telecommunications
关键词:
声音分类学习知觉磁效应自组织特征映射
Keywords:
sound category learningperceptual magnet effectself-organizing feature map
分类号:
TP31
文献标志码:
A
摘要:
语音感知能力是人类感知系统的一个重要方面。在婴儿时期,随着婴儿对母语音位的不断接触和学习,会产生分类感知和知觉磁效应现象,这与自组织特征映射的经典表述相矛盾。文中提出了一种能对这种现象进行解释的神经模型,并通过模型的仿真实验证明了该模型的可行性和正确性。模型进一步预测磁效应的产生是因为原型元音比非原型元音有更小的听觉皮层表示,通过功能磁共振成像实验对模型提出的预测进行了论证
Abstract:
Speech perception is an important part of the perception system. The category learning during infancy leads to the phenomena such as categorical perception and the perceptual magnet effect,which conflicts with the classical formulation of a self-organizing feature map. It proposes a model to explain this phenomenon and the simulation of model proves the feasibility and correctness of this model. The model further predicts that the magnet effect arises because prototypical vowels have smaller auditory cortical representations than nonprototypical vowels, which was supported by a functional magnetic resonance imaging ( fMRI ) experiment

备注/Memo

备注/Memo:
国家自然科学基金(61073115)王莉莉(1989-),女,江西九江人,硕士研究生,研究方向为模式识别与智能系统;张少白,硕士研究生导师,研究方向为人工智能与认知科学,信息获取、处理与识别等
更新日期/Last Update: 1900-01-01