图像搜索中人机交互技术的新进展  

Human-computer interaction technology in image searches: a survey

作　　者：张菁[1] 沈兰荪[1] David Dagan FENG 

机构地区：[1]北京工业大学信号与信息处理研究室,北京100022 [2]悉尼大学信息学院

出　　处：《智能系统学报》2007年第4期14-20,共7页CAAI Transactions on Intelligent Systems

基　　金：国家自然科学基金资助项目(60472036;60431020;60402036);教育部博士点基金资助项目(20040005015);北京市自然科学基金资助项目(3052005);the Poly U/UGC grants(B-Q698)

摘　　要：人机交互在图像搜索中起着重要的作用.研究下一代人机交互接口以更好地表达用户搜索意图,具有广大的应用前景.如何充分利用人类的感觉器官,提供拟人化的交互方式已经成为信息科学的一个研究热点.除了提供自然友好的人机交互,还需要研究如何采用相关反馈技术获取用户的真实需求,以弥补图像底层特征和高层语义之间的鸿沟,优化查询结果,实现个性化搜索.首先对图像搜索的发展概况做了简要介绍,在对人机交互、相关反馈和个性化搜索的研究进展进行讨论后,描述了人眼跟踪、语音和触摸导航在图像检索中的应用.最后指出了图像搜索中人机交互技术进一步的发展前景.Human-computer interaction computer interactions which can identi plays an important role in image fy users＇ search intentions are a searches. Next generation human promising research field. Ways to do this by fully utilizing human sense organs and providing human-like interaction have become a lively topic in informatics. Based on a natural and friendly human-computer interaction, relevance feedback is used to determine a user＇s requirements and narrow the gap between low-level image features and high-level semantic concepts in order to optimize query results and perform a personalized search. Developments in the area of image searches are briefly addressed. The current state of human-computer interaction, relevance feedback, and speech, and haptical personalized search navigation are also are discussed. A described. Fina ppiications for image retrieval using eye-tracking, Finally current challenges and future trends are outlined.

关 键 词：人机交互 图像搜索 相关反馈 语义鸿沟 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]