Bootstrapping deep feature hierarchy for pornographic image recognition

CASIA OpenIR > 多模态人工智能系统全国重点实验室 > 视频内容安全

	Bootstrapping deep feature hierarchy for pornographic image recognition
	Kai Li1 ; Junliang Xing1 ; Bing Li1 ; Weiming Hu1,2
	2016
会议名称	IEEE International Conference on Image Processing
页码	4423-4427
会议日期	September 25 - 28, 2016
会议地点	Arizona, USA
摘要	Automatically recognizing pornographic images from the Web is a vital step to purify Internet environment. Inspired by the rapid developments of deep learning models, we present a deep architecture of convolutional neural network (CNN) for high accuracy pornographic image recognition. The proposed architecture is built upon existing CNNs which accepts input images of different sizes and incorporates features from different hierarchy to perform prediction. To effectively train the model, we propose a two-stage training strategy to learn the model parameters from scratch and end-to-end. During the training procedure, we also employ a hard negative sampling strategy to further reduce the false positive rate of the model. Experimental results on a large dataset demonstrate good performance of the proposed model and the effectiveness of our training strategies, with a considerable improvement over some traditional methods using hand-crafted features and deep learning method using mainstream CNN architecture.
关键词	Pornographic Image Recognition Deep Learning Bootstrap
收录类别	EI
语种	英语
文献类型	会议论文
条目标识符	http://ir.ia.ac.cn/handle/173211/13481
专题	多模态人工智能系统全国重点实验室_视频内容安全
作者单位	1.National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, P. R. China 2.CAS Center for Excellence in Brain Science and Intelligence Technology, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, P. R. China
第一作者单位	模式识别国家重点实验室
推荐引用方式 GB/T 7714	Kai Li,Junliang Xing,Bing Li,et al. Bootstrapping deep feature hierarchy for pornographic image recognition[C],2016:4423-4427.

条目包含的文件		下载所有文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
ICIP16BootstappingDe（245KB）	会议论文		开放获取	CC BY-NC-SA	浏览下载