CASIA OpenIR

Browse/Search Results:  1-10 of 721 Help

Selected(0)Clear Items/Page:    Sort:
基于多域听觉特征建模的说话人无关语音分离方法研究 学位论文
, 北京市: 中国科学院自动化研究所, 2021
Authors:  范存航
Adobe PDF(3377Kb)  |  Favorite  |  View/Download:15/1  |  Submit date:2021/06/01
说话人无关语音分离  听觉特征建模  深度嵌入式特征  深度注意力融合特征  门控递归融合  
基于视觉构图建模的图像编辑问题研究 学位论文
工学博士, 中国科学院自动化研究所: 中国科学院大学, 2021
Authors:  李德榜
Adobe PDF(43095Kb)  |  Favorite  |  View/Download:16/2  |  Submit date:2021/05/31
视觉构图建模  图像编辑  视觉美感  图像裁剪  图像子区域推荐  
仿生滑翔机器海豚的多模态运动控制研究 学位论文
, 北京: 中国科学院大学, 2021
Authors:  王健
Adobe PDF(8779Kb)  |  Favorite  |  View/Download:19/3  |  Submit date:2021/06/04
仿生滑翔机器海豚  多模态  深度控制  路径规划  路径跟踪  
基于受限样本的图像转换方法研究 学位论文
, 中国科学院自动化研究所: 中国科学院大学, 2021
Authors:  曹杰
Adobe PDF(12921Kb)  |  Favorite  |  View/Download:13/1  |  Submit date:2021/05/31
图像转换  图像合成  生成对抗式网络  人脸转正  异质人脸图像生成  
Decomposed Meta Batch Normalization for Fast Domain Adaptation in Face Recognition 期刊论文
IEEE Transactions on Information Forensics and Security, 2021, 卷号: 16, 期号: -, 页码: 3082-3095
Authors:  Guo JZ(郭建珠);  Zhu XY(朱翔昱);  Lei Z(雷震);  Li ZQ(李子青)
Adobe PDF(4808Kb)  |  Favorite  |  View/Download:1/1  |  Submit date:2021/05/31
Face recognition  unsupervised domain adaptation  meta-learning  batch normalization  
Joint stroke classification and text line grouping in online handwritten documents with edge pooling attention networks 期刊论文
Pattern Recognition, 2021, 卷号: 114, 期号: 114, 页码: 107859
Authors:  Jun-Yu Ye;  Yan-Ming Zhang;  Qing Yang;  Cheng-Lin Liu
Adobe PDF(1780Kb)  |  Favorite  |  View/Download:42/16  |  Submit date:2021/03/30
Online handwritten documents  Stroke classification  Text line grouping  Graph neural networks  Edge pooling attention networks  
2D and 3D Palmprint and Palm Vein Recognition Based on Neural Architecture Search 期刊论文
International Journal of Automation and Computing, 2021, 卷号: 18, 期号: 3, 页码: 377-409
Authors:  Wei Jia;  Wei Xia;  Yang Zhao;  Hai Min;  Yan-Xiang Chen
Adobe PDF(15758Kb)  |  Favorite  |  View/Download:4/0  |  Submit date:2021/05/24
Performance evaluation  neural architecture search  biometrics  palmprint  palm vein  deep learning  
EDT Method for Multiple Labelled Objects Subject to Tied Distances 期刊论文
International Journal of Automation and Computing, 2021, 卷号: 18, 期号: 3, 页码: 468-479
Authors:  Andre Marasca;  Andre Backes;  Fabio Favarim;  Marcelo Teixeira;  Dalcimar Casanova
Adobe PDF(2313Kb)  |  Favorite  |  View/Download:7/1  |  Submit date:2021/05/24
Euclidean distance transform (EDT)  multiple-labelled objects  tied distances  fractal analysis  texture analysis  
Deep Audio-Visual Learning: A Survey 期刊论文
International Journal of Automation and Computing, 2021, 卷号: 18, 期号: 3, 页码: 351-376
Authors:  Hao Zhu;  Man-Di Luo;  Rui Wang;  Ai-Hua Zheng;  Ran He
Adobe PDF(1864Kb)  |  Favorite  |  View/Download:4/0  |  Submit date:2021/05/24
Deep audio-visual learning  audio-visual separation and localization  correspondence learning  generative models  representation learning  
PokerNet: Expanding Features Cheaply via Depthwise Convolutions 期刊论文
International Journal of Automation and Computing, 2021, 卷号: 18, 期号: 3, 页码: 432-442
Authors:  Wei Tang;  Yan Huang;  Liang Wang
Adobe PDF(1163Kb)  |  Favorite  |  View/Download:3/0  |  Submit date:2021/05/24
Deep learning  depthwise convolution  lightweight deep model  model compression  model acceleration