期刊
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
卷 13, 期 2, 页码 520-531出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TII.2016.2605629
关键词
Content based; in-video ads; object level optimization; video advertising
类别
资金
- Natural Science Foundation of China [61300209, 61572156]
- Shenzhen Science and Technology Program [JCYJ20150625142543464, JSGG 20150512145714247, JCYJ20160330163900579]
In this paper, we present new models and algorithms for object-level video advertising. A framework that aims to embed content-relevant ads within a video stream is investigated in this context. First, a comprehensive optimization model is designed to minimize intrusiveness to viewers when ads are inserted in a video. For human clothing advertising, we design a deep convolutional neural network using face features to recognize human genders in a video stream. Human parts alignment is then implemented to extract human part features that are used for clothing retrieval. Second, we develop a heuristic algorithm to solve the proposed optimization problem. For comparison, we also employ the genetic algorithm to find solutions approaching the global optimum. Our novel framework is examined in various types of videos. Experimental results demonstrate the effectiveness of the proposed method for object-level video advertising.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据