CASIA OpenIR

浏览/检索结果: 共6条,第1-6条 帮助

限定条件            
已选(0)清除 条数/页:   排序方式:
Faceboxes: A CPU real-time and accurate unconstrained face detector 期刊论文
NEUROCOMPUTING, 2019, 期号: 364, 页码: 297-309
作者:  Zhang, Shifeng;  Wang, Xiaobo;  Lei, Zhen;  Li, Stan Z.
浏览  |  Adobe PDF(5723Kb)  |  收藏  |  浏览/下载:304/47  |  提交时间:2019/12/16
Face detection  CPU real-time  Convolutional neural network  
Re-KISSME: A robust resampling scheme for distance metric learning in the presence of label noise 期刊论文
NEUROCOMPUTING, 2019, 卷号: 330, 期号: 22, 页码: 138-150
作者:  Zeng, Fanxia;  Zhang, Wensheng;  Zhang, Siheng;  Zheng, Nan
Adobe PDF(917Kb)  |  收藏  |  浏览/下载:392/59  |  提交时间:2019/07/12
Resampling scheme  KISSME  Distance metric learning  Label noise  
Learning view invariant gait features with Two-Stream GAN 期刊论文
NEUROCOMPUTING, 2019, 卷号: 339, 期号: 2019, 页码: 245-254
作者:  Wang, Yanyun;  Song, Chunfeng;  Huang, Yan;  Wang, Zhenyu;  Wang, Liang
浏览  |  Adobe PDF(2222Kb)  |  收藏  |  浏览/下载:508/142  |  提交时间:2019/07/12
Gait recognition  Cross-veiw  Two-Stream GAN  
Reading scene text with fully convolutional sequence modeling 期刊论文
NEUROCOMPUTING, 2019, 卷号: 339, 期号: 无, 页码: 161-170
作者:  Gao, Yunze;  Chen, Yingying;  Wang, Jinqiao;  Tang, Ming;  Lu, Hanqing
浏览  |  Adobe PDF(1725Kb)  |  收藏  |  浏览/下载:387/100  |  提交时间:2019/07/12
Fully convolutional sequence modeling  Scene text recognition  
A hierarchical contextual attention-based network for sequential recommendation 期刊论文
NEUROCOMPUTING, 2019, 卷号: 358, 页码: 141-149
作者:  Cui, Qiang;  Wu, Shu;  Huang, Yan;  Wang, Liang
Adobe PDF(545Kb)  |  收藏  |  浏览/下载:279/28  |  提交时间:2019/05/09
Sequential recommendation  Recurrent neural network  Short-term interest  Context  Attention mechanism  
Rough extreme learning machine: A new classification method based on uncertainty measure 期刊论文
NEUROCOMPUTING, 2019, 卷号: 325, 页码: 269-282
作者:  Feng, Lin;  Xu, Shuliang;  Wang, Feilong;  Liu, Shenglan;  Qiao, Hong
收藏  |  浏览/下载:294/0  |  提交时间:2019/01/08
Extreme learning machine  Rough set  Attribute reduction  Classification  Neural network