论文部分内容阅读
2014年度北京市科学技术奖一等奖获奖项目展示大规模网络视频处理与内容分析关键技术及应用主要完成人:张勇东、张冬明、李建涛、郭俊波、李锦涛、唐胜、曹娟、代锋、高科、顾晓光、王波、马宜科、靳国庆、崔洪亮、李正勇完成单位:中国科学院计算技术研究所、北京北纬通信科技股份有限公司、北京神州天脉网络计算机有限公司该成果的创新点包括:1.发明了一套视频编解码优化及适配处理方法,将单机解码吞吐能力从原有的60Mbps提升到300Mbps,显著提高了网络视频的解码效率。2.发明了一套适用于大规模相似视频检测的特征提取、匹配与高维索引方法,实现了高速、高精度的大规模相似视频检测,解
2014 First Prize of Beijing Science and Technology Award Winning Project Key Technologies and Applications of Massive Network Video Processing and Content Analysis Main Completed by: Zhang Yongdong, Zhang Dongming, Li Jiantao, Guo Junbo, Li Jintao, Tang Sheng, Cao Juan, Dai Feng, Gao Ke, Gu Xiaoguang, Wang Bo, Ma Yike, Jin Guoqing, Cui Hongliang, Li Zhengyong Completed by: Institute of Computing Technology, Chinese Academy of Sciences, Beijing Latitude Communication Technology Co., Ltd., Beijing Shenzhou Tianyi Network Computer Co., Ltd. The innovations in this achievement include: A set of video codec optimization and adaptation processing methods, single-machine decoding throughput from the original 60Mbps to 300Mbps, significantly improve the network video decoding efficiency. 2. A set of feature extraction, matching and high-dimensional indexing methods suitable for large-scale similar video detection has been invented to realize high-speed, high-precision large-scale similar video detection