声学场景深度识别系统设计  

Talk about Acoustic Scene Recognition System

作　　者：李嘉[1] 黄程韦[2] 韩辰[1] 

机构地区：[1]江苏省广播电视总台,南京210013 [2]苏州大学,苏州215006

出　　处：《数字通信世界》2016年第7期12-14,共3页Digital Communication World

摘　　要：基于视频的监控系统存在很多不足,声学场景识别系统依据基于人耳听觉仿生的目标声音识别技术,积极探索先进的人耳仿生理论、特征提取技术、目标声音分类技术,实现对声音场景进行自动化分类和信息抽取,具有广泛的应用前景。As video monitoring systems has many deficiencies, Acoustic Scene Recognition System is built on target voice recognition technology based on Bionic ear hearing, and actively explore advanced ear bionic theory, feature extraction, target sound classification technology. It can achieve the sound scene automated classification and information extraction, and has broad application prospects.

关 键 词：声音识别 声音场景 特征提取 神经网络 

分 类 号：TN912[电子电信—通信与信息系统]