Web Objectionable Video Recognition Based on Deep Multi-Instance Learning With Representative Prototypes Selection

Web Objectionable Video Recognition Based on Deep Multi-Instance Learning With Representative Prototypes Selection
复制标题

基于深度多实例学习和代表性原型选择的网络不良视频识别

DOI:
10.1109/tcsvt.2020.2992276
复制
发表时间:
2021-03
影响因子:
8.4
通讯作者:
Weiming Hu
Weiming Hu
中科院分区:
工程技术1区
文献类型:
--
作者:
Xinmiao Ding;Bing Li;Yangxi Li;Wen Guo;Yao Liu;Weihua Xiong;Weiming Hu

文献摘要

参考文献

相似文献

为了防止未成年人访问互联网上的不良视频,需要一种有效的不良视频识别算法来进行网络过滤。最近,多实例学习被引入到令人反感的视频识别中,并取得了令人印象深刻的结果。然而,手工制作的特征以及令人反感的视频中的冗余和噪声帧成为一个棘手的问题,不可避免地降低识别性能。在本文中,我们提出了一种嵌入深度多实例表示学习的新颖的代表性原型选择算法。在该方法中,设计了一种用于多模态多实例特征学习的改进卷积神经网络,并设计了一种基于稀疏和低秩约束的自表达字典学习模型,以从实例的每个子空间中选择代表性原型。然后通过将包映射到选定的原型来构建包级特征。对三个不良视频集的实验表明了我们的方法对于不良视频识别的有效性。
To protect underage people from accessing objectionable videos in the Internet, an effective objectionable video recognition algorithm is necessary for web filtering. Recently, the multi-instance learning has been introduced for objectionable video recognition and achieves impressive results. However, hand-crafted features as well as redundant and noisy frames in objectionable videos become an intractable problem that inevitably degrades the recognition performance. In this paper, we propose a novel representative prototype selection algorithm embedding deep multi-instance representation learning. In the proposed method, an improved convolutional neural network is designed for multimodal multi-instance feature learning and a self-expressive dictionary learning model based on sparse and low rank constraint is designed to select the representative prototypes from each subspace of instances. Then the bag-level feature is constructed via mapping the bag to the selected prototypes. Experiments on three objectionable video sets show the effectiveness of our method for objectionable video recognition.
DOI: 10.1007/978-3-642-42057-3_57
发表时间: 2013-07
影响因子: 4.7
作者:
Shuai Hao;Ou Wu;Weiming Hu;Jinfeng Yang
通讯作者: Shuai Hao;Ou Wu;Weiming Hu;Jinfeng Yang
DOI: 10.1109/tpami.2018.2874455
发表时间: 2020-01-01
影响因子: 23.6
作者:
Ye, Jun;Qi, Guo-Jun;Hua, Kien A.
通讯作者: Hua, Kien A.
DOI: 10.1109/tip.2016.2612825
发表时间: 2017-01
期刊: IEEE transactions on image processing : a publication of the IEEE Signal Processing Society
影响因子: --
作者:
Jinhui Tang;Shiyu Chang;Guo-jun Qi;Qi Tian;Yong Rui;Thomas S. Huang
通讯作者: Jinhui Tang;Shiyu Chang;Guo-jun Qi;Qi Tian;Yong Rui;Thomas S. Huang
DOI: 10.1109/tpami.2007.1133
发表时间: 2007-06
影响因子: 23.6
作者:
Weiming Hu;Ou Wu;Zhouyao Chen;Zhouyu Fu;S. Maybank
通讯作者: Weiming Hu;Ou Wu;Zhouyao Chen;Zhouyu Fu;S. Maybank
DOI: 10.1109/tcsvt.2005.856896
发表时间: 2006-01-01
影响因子: 8.4
作者:
Cerneková, Z;Pitas, I;Nikou, C
通讯作者: Nikou, C