检索结果-内蒙古大学图书馆

arXiv 2019年

作者： Wan, Ruosi Xiong, Haoyi Li, Xingjian Zhu, Zhanxing Huan, Jun Big Data Laboratory Baidu Inc. Beijing China National Engineering Laboratory for Deep Learning Technology and Applications Beijing China School of Mathematical Sciences Peking University Beijing China

—Transfer learning have been frequently used to improve deep neural network training through incorporating weights of pre-trained networks as the starting-point of optimization for regularization. While deep transfer learning can usually boost the performance with better accuracy and faster convergence, transferring weights from inappropriate networks hurts training procedure and may lead to even lower accuracy. In this paper, we consider deep transfer learning as minimizing a linear combination of empirical loss and regularizer based on pre-trained weights, where the regularizer would restrict the training procedure from lowering the empirical loss, with conflicted descent directions (e.g., derivatives). Following the view, we propose a novel strategy making regularization-based deep Transfer learning Never Hurt (DTNH) that, for each iteration of training procedure, computes the derivatives of the two terms separately, then re-estimates a new descent direction that does not hurt the empirical loss minimization while preserving the regularization affects from the pre-trained weights. Extensive experiments have been done using common transfer learning regularizers, such as L2-SP and knowledge distillation, on top of a wide range of deep transfer learning benchmarks including Caltech, MIT indoor 67, CIFAR-10 and ImageNet. The empirical results show that the proposed descent direction estimation strategy DTNH can always improve the performance of deep transfer learning tasks based on all above regularizers, even when transferring pre-trained weights from inappropriate networks. All in all, DTNH strategy can improve state-of-the-art regularizers in all cases with 0.1%—7% higher accuracy in all experiments. Copyright © 2019, The Authors. All rights reserved.

关键词： deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Q-ViT: accurate and fully quantized low-bit vision transformer 22

Q-ViT: accurate and fully quantized low-bit vision transform...

引用

Proceedings of the 36th International Conference on Neural Information Processing Systems

作者： Yanjing Li Sheng Xu Baochang Zhang Xianbin Cao Peng Gao Guodong Guo Beihang University Beijing P.R.China Beihang University Beijing P.R.China and Zhongguancun Laboratory Beijing P.R.China Shanghai Artificial Intelligence Laboratory Shanghai P.R.China Institute of Deep Learning Baidu Research Beijing P.R.China and National Engineering Laboratory for Deep Learning Technology and Application Beijing P.R.China

ISBN: (纸本)9781713871088

The large pre-trained vision transformers (ViTs) have demonstrated remarkable performance on various visual tasks, but suffer from expensive computational and memory cost problems when deployed on resource-constrained devices. Among the powerful compression approaches, quantization extremely reduces the computation and memory consumption by low-bit parameters and bit-wise operations. However, low-bit ViTs remain largely unexplored and usually suffer from a significant performance drop compared with the real-valued counterparts. In this work, through extensive empirical analysis, we first identify the bottleneck for severe performance drop comes from the information distortion of the low-bit quantized self-attention map. We then develop an information rectification module (IRM) and a distribution guided distillation (DGD) scheme for fully quantized vision transformers (Q-ViT) to effectively eliminate such distortion, leading to a fully quantized ViTs. We evaluate our methods on popular DeiT and Swin backbones. Extensive experimental results show that our method achieves a much better performance than the prior arts. For example, our Q-ViT can theoretically accelerates the ViT-S by 6.14 × and achieves about 80.9% Top-1 accuracy, even surpassing the full-precision counterpart by 1.0% on ImageNet dataset.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A unified object motion and affinity model for online Multi-Object Tracking

arXiv

引用

arXiv 2020年

作者： Yin, Junbo Wang, Wenguan Meng, Qinghao Yang, Ruigang Shen, Jianbing Beijing Lab of Intelligent Information Technology School of Computer Science Beijing Institute of Technology China ETH Zurich Switzerland Baidu Research China National Engineering Laboratory of Deep Learning Technology and Application China Inception Institute of Artificial Intelligence United Arab Emirates University of Kentucky Kentucky United States

Current popular online multi-object tracking (MOT) solutions apply single object trackers (SOTs) to capture object motions, while often requiring an extra affinity network to associate objects, especially for the occluded ones. This brings extra computational overhead due to repetitive feature extraction for SOT and affinity computation. Meanwhile, the model size of the sophisticated affinity network is usually non-trivial. In this paper, we propose a novel MOT framework that unifies object motion and affinity model into a single network, named UMA, in order to learn a compact feature that is discriminative for both object motion and affinity measure. In particular, UMA integrates single object tracking and metric learning into a unified triplet network by means of multi-task learning. Such design brings advantages of improved computation efficiency, low memory requirement and simplified training procedure. In addition, we equip our model with a task-specific attention module, which is used to boost task-aware feature learning. The proposed UMA can be easily trained end-to-end, and is elegant – requiring only one training stage. Experimental results show that it achieves promising performance on several MOT Challenge benchmarks. Copyright © 2020, The Authors. All rights reserved.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

deeply supervised depth map super-resolution as novel view synthesis

arXiv

引用

arXiv 2018年

作者： Song, Xibin Dai, Yuchao Qin, Xueying School of Computer Science and Technology Shandong University China Baidu Research Beijing China National Engineering Laboratory of Deep Learning Technology and Application China Shaanxi Key Lab of Information Acquisition and Processing School of Electronics and Information Northwestern Polytechnical University School of Software Shandong University China

deep convolutional neural network (DCNN) has been successfully applied to depth map super-resolution and outperforms existing methods by a wide margin. However, there still exist two major issues with these DCNN based depth map super-resolution methods that hinder the performance: i) The low-resolution depth maps either need to be up-sampled before feeding into the network or substantial deconvolution has to be used;and ii) The supervision (high-resolution depth maps) is only applied at the end of the network, thus it is difficult to handle large up-sampling factors, such as ×8;×16. In this paper, we propose a new framework to tackle the above problems. First, we propose to represent the task of depth map super-resolution as a series of novel view synthesis sub-tasks. The novel view synthesis sub-task aims at generating (synthesizing) a depth map from different camera pose, which could be learned in parallel. Second, to handle large up-sampling factors, we present a deeply supervised network structure to enforce strong supervision in each stage of the network. Third, a multi-scale fusion strategy is proposed to effectively exploit the feature maps at different scales and handle the blocking effect. In this way, our proposed framework could deal with challenging depth map super-resolution efficiently under large up-sampling factors (e.g.×8;×16). Our method only uses the low-resolution depth map as input, and the support of color image is not needed, which greatly reduces the restriction of our method. Extensive experiments on various benchmarking datasets demonstrate the superiority of our method over current state-of-the-art depth map super-resolution methods. Copyright © 2018, The Authors. All rights reserved.

关键词： Optical resolving power

来源：评论

学校读者我要写书评

暂无评论

Relaxed 2-D principal component analysis by Lpnorm for face recognition

arXiv

引用

arXiv 2019年

作者： Chen, Xiao Jia, Zhi-Gang Cai, Yunfeng Zhao, Mei-Xiang School of Mathematics and Statistics Jiangsu Key Laboratory of Education Big Data Science and Engineering Jiangsu Normal University Xuzhou221116 China Baidu Research National Engineering Laboratory for Deep Learning Technology and Applications Beijing100193 China

A relaxed two dimensional principal component analysis (R2DPCA) approach is proposed for face recognition. Different to the 2DPCA, 2DPCA-L1 and G2DPCA, the R2DPCA utilizes the label information (if known) of training samples to calculate a relaxation vector and presents a weight to each subset of training data. A new relaxed scatter matrix is defined and the computed projection axes are able to increase the accuracy of face recognition. The optimal Lp-norms are selected in a reasonable range. Numerical experiments on practical face databased indicate that the R2DPCA has high generalization ability and can achieve a higher recognition rate than state-of-the-art methods. Copyright © 2019, The Authors. All rights reserved.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

A Spatio-Spectral Hybrid Convolutional Architecture for Hyperspectral Document Authentication

A Spatio-Spectral Hybrid Convolutional Architecture for Hype...

引用

International Conference on Document Analysis and Recognition

作者： Muhammad Jaleed Khan Khurram Khurshid Faisal Shafait Department of Electrical Engineering Institute of Space Technology (IST) Islamabad Pakistan School of Electrical Engineering and Computer Science (SEECS) National University of Sciences and Technology (NUST) Islamabad Pakistan Deep Learning Laboratory National Center of Artificial Intelligence (NCAI) Islamabad Pakistan

Hyperspectral Document Image (HSDI) analysis allows for efficient and accurate differentiation of inks with visually similar color but unique spectral response, which is a crucial step in authentication of documents. Various HSDI based ink discrimination methods are available in the current literature, however, more accurate and robust methods are required to empower document authentication. Contrary to the former ink mismatch detection methods based on spectral features only, we present a novel method based on deep learning that exploits the spectral correlation as well as the spatial context to enhance ink mismatch detection. Spectral responses of the target pixel and its neighboring pixels are organized in an image format and fed to a Convolutional Neural Network (CNN) for classification. The proposed method achieves the highest accuracy among the other ink mismatch detection methods on the UWA Writing Ink Hyperspectral Images database (WIHSI), which demonstrates the effectiveness of deep learning models employing spatio-spectral hybrid features for document authentication. Detailed experimental analysis for selection of appropriate CNN architecture, spatio-spectral data format and training ratio is presented along with a comparison with the previous methods on this subject.

关键词： Ink Hyperspectral imaging Feature extraction Authentication deep learning Databases Text analysis

来源：评论

学校读者我要写书评

暂无评论

Nested Collaborative learning for Long-Tailed Visual Recognition

arXiv

引用

arXiv 2022年

作者： Li, Jun Tan, Zichang Wan, Jun Lei, Zhen Guo, Guodong CBSR&NLPR Institute of Automation Chinese Academy of Sciences Beijing China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing China Institute of Deep Learning Baidu Research Beijing China National Engineering Laboratory for Deep Learning Technology and Application Beijing China Centre for Artificial Intelligence and Robotics Hong Kong Institute of Science&Innovation Chinese Academy of Sciences Hong Kong

The networks trained on the long-tailed dataset vary remarkably, despite the same training settings, which shows the great uncertainty in long-tailed learning. To alleviate the uncertainty, we propose a Nested Collaborative learning (NCL), which tackles the problem by collaboratively learning multiple experts together. NCL consists of two core components, namely Nested Individual learning (NIL) and Nested Balanced Online Distillation (NBOD), which focus on the individual supervised learning for each single expert and the knowledge transferring among multiple experts, respectively. To learn representations more thoroughly, both NIL and NBOD are formulated in a nested way, in which the learning is conducted on not just all categories from a full perspective but some hard categories from a partial perspective. Regarding the learning in the partial perspective, we specifically select the negative categories with high predicted scores as the hard categories by using a proposed Hard Category Mining (HCM). In the NCL, the learning from two perspectives is nested, highly related and complementary, and helps the network to capture not only global and robust features but also meticulous distinguishing ability. Moreover, self-supervision is further utilized for feature enhancement. Extensive experiments manifest the superiority of our method with outperforming the state-of-the-art whether by using a single model or an ensemble. Code is available at https://***/Bazinga699/NCL Copyright © 2022, The Authors. All rights reserved.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

Cross-ethnicity face anti-spoofing recognition challenge: A review

arXiv

引用

arXiv 2020年

作者： Liu, Ajian Li, Xuan Wan, Jun Liang, Yanyan Escalera, Sergio Escalante, Hugo Jair Madadi, Meysam Jin, Yi Wu, Zhuoyuan Yu, Xiaogang Tan, Zichang Yuan, Qi Yang, Ruikun Zhou, Benjia Guo, Guodong Li, Stan Z. Faculty of Information Technology Avenida WaiLong Taipa Macau China School of Computer and Information Technology Beijing Jiaotong University Beijing China National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Science Beijing China Universitat de Barcelona and Computer Vision Center Barcelona Instituto Nacional de Astrofísica Óptica y Electrónica Puebla Mexico School of Software Beihang University Beijing China Institute of Deep Learning Baidu Research and National Engineering Laboratory for Deep Learning Technology and Application Beijing Westlake University Hangzhou China

Face anti-spoofing is critical to prevent face recognition systems from a security breach. The biometrics community has achieved impressive progress recently due the excellent performance of deep neural networks and the availability of large datasets. Although ethnic bias has been verified to severely affect the performance of face recognition systems, it still remains an open research problem in face anti-spoofing. Recently, a multi-ethnic face anti-spoofing dataset, CASIA-SURF CeFA, has been released with the goal of measuring the ethnic bias. It is the largest up to date cross-ethnicity face anti-spoofing dataset covering 3 ethnicities, 3 modalities, 1;607 subjects, 2D plus 3D attack types, and the first dataset including explicit ethnic labels among the recently released datasets for face *** organized the Chalearn Face Anti-spoofing Attack Detection Challenge which consists of single-modal (e.g., RGB) and multi-modal (e.g., RGB, Depth, Infrared (IR)) tracks around this novel resource to boost research aiming to alleviate the ethnic bias. Both tracks have attracted 340 teams in the development stage, and finally 11 and 8 teams have submitted their codes in the single-modal and multi-modal face anti-spoofing recognition challenges, respectively. All the results were verified and re-ran by the organizing team, and the results were used for the final ranking. This paper presents an overview of the challenge, including its design, evaluation protocol and a summary of results. We analyze the top ranked solutions and draw conclusions derived from the competition. In addition we outline future work directions. Copyright © 2020, The Authors. All rights reserved.

关键词： Large dataset

来源：评论

学校读者我要写书评

暂无评论

$\mathsf{NCF}$NCF: A Neural Context Fusion Approach to Raw Mobility Annotation

引用

IEEE Transactions on Mobile Computing 2020年第1期21卷 226-238页

作者： Renjun Hu Jingbo Zhou Xinjiang Lu Hengshu Zhu Shuai Ma Hui Xiong SKLSDE Lab Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing China Business Intelligence Lab Baidu Research National Engineering Laboratory of Deep Learning Technology and Application Beijing China Talent Intelligence Center Baidu Inc. Beijing China Management Science and Information Systems Department Rutgers Business School Rutgers University Newark NJ USA

Understanding human mobility patterns at the point-of-interest (POI) scale plays an important role in enhancing business intelligence in mobile environments. While large efforts have been made in this direction, most studies simply utilize POI check-ins to mine the concerned mobility patterns, the effectiveness of which is usually hindered due to data sparsity. To obtain better POI-based human mobility for mining, in this paper, we strive to directly annotate the POIs associated with raw user-generated mobility records. We propose a neural context fusion approach which integrates various context factors in people's POI-visiting behaviors. Our approach evaluates the preference and transition factors via representation learning. Notably, we incorporate an attention mechanism to deal with the randomized transitions in raw mobility. The domain knowledge factors, i.e., distance, time and popularity, remain effective and our approach further includes them from a data-driven perspective. Factors are automatically fused with a feed-forward neural network. Furthermore, we exploit a multi-head architecture to enhance the model expressiveness. Using two real-life data sets, we conduct our experimental study and find that our approach consistently outperforms the state-of-the-art baselines by at least 32 percent in accuracy. Besides, we demonstrate the utility of the obtained POI-based human mobility with a POI recommendation example.

关键词： Annotations Trajectory Mobile computing Computer architecture Business intelligence Neural networks Semantics

来源：评论

学校读者我要写书评

暂无评论

Hyperspectral Image Analysis for Writer Identification using deep learning

Hyperspectral Image Analysis for Writer Identification using...

引用

Proceedings of the Digital Image Computing: Technqiues and applications (DICTA)

作者： Ammad Ul Islam Muhammad Jaleed Khan Khurram Khurshid Faisal Shafait Artificial Intelligence and Computer Vision (iVision) Lab Institute of Space Technology (IST) Islamabad Pakistan School of Electrical Engineering and Computer Science (SEECS) National University of Sciences and Technology (NUST) Islamabad Pakistan Deep Learning Laboratory National Center of Artificial Intelligence (NCAI) Islamabad Pakistan

Handwriting is a behavioral characteristic of human beings that is one of the common idiosyncrasies utilized for litigation purposes. Writer identification is commonly used for forensic examination of questioned and specimen documents. Recent advancements in imaging and machine learning technologies have empowered the development of automated, intelligent and robust writer identification methods. Most of the existing methods based on human defined features and color imaging have limited performance in terms of accuracy and robustness. However, rich spectral information content obtained from hyperspectral imaging (HSI) and suitable spatio-spectral features extracted using deep learning can significantly enhance the performance of writer identification in terms of accuracy and robustness. In this paper, we propose a novel writer identification method in which spectral responses of text pixels in a hyperspectral document image are extracted and are fed to a Convolutional Neural Network (CNN) for writer classification. Different CNN architectures, hyperparameters, spatio-spectral formats, train-test ratios and inks are used to evaluate the performance of the proposed system on the UWA Writing Inks Hyperspectral Images (WIHSI) database and to select the most suitable set of parameters for writer identification. The findings of this work have opened a new arena in forensic document analysis for writer identification using HSI and deep learning.

关键词： Hyperspectral imaging Feature extraction Ink Text analysis Forensics deep learning Databases

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：