英语翻译At query time,the user has option of submitting query in any of the three paradigms as mentioned in section V.For QBE,the user submits an example video clip to the system from which audio-visual feature are extracted for matching in the i

英语翻译At query time,the user has option of submitting query in any of the three paradigms as mentioned in section V.For QBE,the user submits an example video clip to the system from which audio-visual feature are extracted for matching in the i
英语翻译
At query time,the user has option of submitting query in any of the three paradigms as mentioned in section V.For QBE,the user submits an example video clip to the system from which audio-visual feature are extracted for matching in the indexed database.In case of QBS,the user is provided with an interface to sketch the objects in video clip and/or their motion trajectories.For QBK search systems,the keywords entered by user are sent to a knowledge base,which interprets the meaning of each keyword and figures out the approximate shape,color and motion of objects the user is interested in.This information is then mapped into audio-visual features.Once the audio-visual features have been compiled,the knowledge base is consulted which generates the metadata corresponding to these features.The query processor then takes up these features as well as the metadata and searches for similar feature vectors in indexed database with appropriate weights being assigned to different features,producing ranked results corresponding to user query,which are finally displayed to the user.
The main hallmarks of this system are that it outlines a system capable of exploiting the object-based representation incorporated in MPEG-4 compressed bitstream; integrates research efforts from many different disciplines and uses multimodal features for representation of video clip.The knowledge base module depicts semantic-level representation from multimedia features.The authors are of the view that a system with semantic representation capabilities,operating on multimodal feature representation of the video content is bound to generate promising results.

英语翻译At query time,the user has option of submitting query in any of the three paradigms as mentioned in section V.For QBE,the user submits an example video clip to the system from which audio-visual feature are extracted for matching in the i
在提问的时候,用户选择提出的质疑的三个范例中提及的部分,用户对QBE弧菌提交实例视频剪辑的系统特征提取,视听匹配在索引数据库.万一QBS,用户的界面来描绘物体在视频剪辑和/或他们的运动轨迹.搜索系统,为QBK关键词进入由用户发送到一个知识库的意义,对每个关键字和数据的近似造型、色彩和运动物体的用户感兴趣的东西.这个信息是映射到影音功能.一旦视听特征已经被编译、知识库咨询产生的元数据对应这些特征.处理器然后占用的质疑这些特征以及寻找相似的数据和特征向量的索引数据库与适当的重量被指派到不同的特点,对用户查询的结果相对应的排名,最终被显示在用户.
该系统的主要标志就是它概述了一个系统的开发能力的表现纳入object-based;集压缩位mpeg - 4的努力从许多不同学科的特点,采用复合视频剪辑的表现.知识库模块表示多媒体功能semantic-level刻画.作者认为一个系统,操作语义表达能力的通道特征表现的视频内容必然产生好的结果.