检索结果-内蒙古大学图书馆

2020 3rd International Conference on Computer Information Science and Artificial Intelligence, CISAI 2020

作者： Chen, Tianyuan Wang, Hongfei Liu, Hao Wu, Peng Key Laboratory of Space Utilization Technology and Engineering Center for Space Utilization Chinese Academy of Sciences Beijing100094 China University of Chinese Academy of Sciences Beijing100049 China State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing100049 China

With the increasing importance of islands in many fields, it has become the focus of research to obtain information from island remote sensing images efficiently by using image semantic segmentation algorithm. In recent years, deep learning methods based on convolutional neural network have been widely used in image segmentation. However, in view of the problems that remote sensing images contain richer ratio information and complex background interference, we propose an island remote sensing image segmentation algorithm based on a fusion network with attention mechanism, called AFU-Net. The network is built on the basis of FC_U-Net [1]. An attention mechanism is added to pre-weight the shallow features before deep-shallow layer feature fusion, in order to enhance the response capability of the target features, suppress the background interference and improve the segmentation accuracy of the network. The testing and comparative experiments on NWPU-RESISC45 dataset show that the quantitative metrics and visual effects of AFU-Net are greatly improved compared to FC_U-Net, and are also superior to other three state-of-the-art methods, U-Net, FCN, SegNet, which indicates the effectiveness of our method. © Published under licence by IOP Publishing Ltd.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

An Island Remote Sensing Image Segmentation Algorithm Based on FC_U-Net Network

An Island Remote Sensing Image Segmentation Algorithm Based ...

引用

2020 International Conference on Computer Vision, Image and Deep Learning, CVIDL 2020

作者： Chen, Tianyuan Wang, Hongfei Liu, Hao Du, Junrong Wu, Peng Key Laboratory of Space Utilization Technology and Engineering Center for Space Utilization Chinese Academy of Sciences Beijing100094 China University of Chinese Academy of Sciences Beijing100049 China State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing100049 China

ISBN: (数字)9781728194813

ISBN: (纸本)9781728194813

With the increasing importance of islands in many fields, it has become the focus of research to obtain information from island remote sensing images efficiently by using image semantic segmentation algorithm. In recent years, deep learning methods based on convolutional neural network have been widely used in image segmentation. However, for remote sensing images with rich details and different target scales, there are still some problems such as insufficient segmentation accuracy. To handle these problems, we propose an island remote sensing image segmentation algorithm based on FC_U-Net network. In this network, U-Net and FCN (Fully Convolutional Network) are improved respectively to form two new modules, and then the semantic features and detail features from the two modules are fused by a fusion module. The testing and comparative experiments on NWPU-RESISC45 dataset show that the quantitative metrics and visual effects of FC_U-Net are better than other three state-of-the-art methods, U-Net, FCN (Fully Convolutional Network), SegNet, which indicates the effectiveness of our method. © 2020 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Robotic Autonomous Grasping Technique: A Survey

Robotic Autonomous Grasping Technique: A Survey

引用

Artificial Intelligence Technology (ACAIT), Asian Conference on

作者： Lili Wang Zhen Zhang Jianhua Su Qipeng Gu the School of Artificial Intelligence University of Chinese Academy of Sciences Beijing P.R. China the State Key Lab. of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing P.R. China

ISBN: (纸本)9781665426312

This paper provides a comprehensive survey of robotic autonomous grasping techniques. We summarize three key tasks: grasp detection, affordance detection, and model migration. Grasp detection determines the graspable area and grasping posture of the manipulator, so that the robot can successfully perform the grasps. The grasp detection methods based on deep learning are divided into 3DoF grasp and 6DoF grasp. The object affordances based grasping methods can further improve the robot's understanding of objects and environment, thereby improving the robot's intelligence and autonomy. Methods for object affordances detection are classified as learning-based, knowledge-based, and simulation-based. Model migration means that when the grasping model is migrated to other scenes where lightness and background changes, only little or no label data is required, so that the grasping model can be used in the target scene quickly and efficiently. This paper focuses on domain adaptation (DA) methods in model migration.

关键词： Deep learning Adaptation models Affordances Knowledge based systems Focusing Grasping Manipulators

来源：评论

学校读者我要写书评

暂无评论

Glaucoma Detection: Joint Segmentation and Classification Framework via Deep Ensemble Network

Glaucoma Detection: Joint Segmentation and Classification Fr...

引用

International Conference on Advanced Robotics and Mechatronics (ICARM)

作者： Siyu Li Zhen Li Limin Guo Gui-Bin Bian Beijing University of Technology Beijing PR China State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China

ISBN: (数字)9781728164793

ISBN: (纸本)9781728164809

Clinical research shows that glaucoma is primarily caused by pathological changes in the optic nerve structure, which may bring about irreversible damage of sight. In relevant literature research, the cup-to-disc ratio (CDR) is mainly used as an important indicator for glaucoma detection, which needs to segment optic disc (OD) and optic cup (OC) region clearly and accurately. However, due to the low contrast image of boundary, the segmentation of OD and OC is still a challenging problem. In this paper, a novel hybrid model based on the ensemble random-forest deep-neural-network (RF-DNN) is proposed for OD and OC segmentation, which can calculate more accurately for glaucoma detection. At the same time, the advantages of DNN and ensemble RF are combined to carry out the corresponding feature extraction and classification, which employs the winner-takes-all strategically to the segmentation and classification for glaucoma detection. Finally, the experiment result shows that the proposed method has reached the best evaluation level in terms of OD and OC segmentation results on ORIGA dataset and SCES dataset, which achieves highest diagnosis accuracy with AUC of 0.96 and 0.98 on ORIGA dataset and SCES dataset, respectively.

关键词： Feature extraction Image segmentation Training Optical imaging Decoding Neural networks Robots

来源：评论

学校读者我要写书评

暂无评论

DeepFatigueNet: A model for automatic visual fatigue assessment based on raw single-channel EEG

DeepFatigueNet: A model for automatic visual fatigue assessm...

引用

SID Symposium, Seminar, and Exhibition 2019, Display Week 2019

作者： Song, Yaguang Wang, Danli Yue, Kang Zheng, Nan The State Key Laboratory of Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781510896161

Three-dimension (3D) display has become increasingly popular in many fields. However, watching 3D content continuously can lead to visual fatigue that is harmful to users’ vision system. Visual fatigue assessment aims at monitoring users’ brain states based on the electroencephalogram (EEG) signals to identify different fatigue levels and avoid severe fatigue. Most of existing studies on the modeling of visual fatigue assessment rely on manual features extracted from EEG, which is time-consuming and needs prior knowledge. Convolutional Neural Networks (CNNs) which have been used in computer vision, speech recognition have attracted increasing interest. There is still a lack of attempts to employ end- to-end EEG analysis on visual fatigue assessment In this paper, we propose a deep learning model DeepFatigueNet to perform automatic feature extraction and classification from raw single-channel EEG. The DeepFatigueNet is evaluated on our own visual fatigue dataset and compared with the state-of-the-art deep learning methods for EEG-based tasks. The overall accuracy of DeepFatigueNet reaches 75.9% on the three-classification task exceeding other models. The experimental results demonstrate the effectiveness of our model and show the potential of deep convolutional neural networks for end-to-end visual fatigue assessment. © 2019 SID-the Society for Information Display. All rights reserved.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

Device Placement Optimization for Deep Neural Networks via One-shot Model and Reinforcement Learning

Device Placement Optimization for Deep Neural Networks via O...

引用

IEEE Symposium Series on Computational Intelligence (SSCI)

作者： Zixiang Ding Yaran Chen Nannan Li Dongbin Zhao State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China

ISBN: (数字)9781728125473

ISBN: (纸本)9781728125480

With the development of deep learning that employs deep neural networks (DNN) as powerful tool, its computational requirement grows rapid together with the increasing size (e.g. depth and parameter) of DNN. Currently, model and data parallelism are employed for accelerating the training and inference process of DNN. However, the above techniques make placement decision on devices for DNN based on heuristics and intuitions by machine learning experts. In this paper, we propose an novel approach for designing device placement of DNN in an automatic way. For a DNN, we employ a sequence-to-sequence model as controller to sample device placement from a one-shot model, which contains all possible device placements with respect to a specific hardware environment (e.g. CPU and GPU). Then, reinforcement learning treats the execution time of sampled device placement as reward to guide the sequence-to-sequence model for finding better one. The proposed approach is employed to optimize the device placement for both model and data parallelism of Inception-V3 on ImageNet. Experimental results show that the optimal placements discovered by our method can outperform hand-crafted one.

关键词： Parallel processing Data models Computational modeling Optimization Reinforcement learning Hardware Training

来源：评论

学校读者我要写书评

暂无评论

DGST: Discriminator Guided Scene Text detector

arXiv

引用

arXiv 2020年

作者： Zhao, Jinyuan Wang, Yanna Xiao, Baihua Shi, Cunzhao Jia, Fuxi Wang, Chunheng State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences China

Scene text detection task has attracted considerable attention in computer vision because of its wide application. In recent years, many researchers have introduced methods of semantic segmentation into the task of scene text detection, and achieved promising results. This paper proposes a detector framework based on the conditional generative adversarial networks to improve the segmentation effect of scene text detection, called DGST (Discriminator Guided Scene Text detector). Instead of binary text score maps generated by some existing semantic segmentation based methods, we generate a multi-scale soft text score map with more information to represent the text position more reasonably, and solve the problem of text pixel adhesion in the process of text extraction. Experiments on standard datasets demonstrate that the proposed DGST brings noticeable gain and outperforms state-of-the-art methods. Specifically, it achieves an F-measure of 87% on ICDAR 2015 dataset. Copyright © 2020, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Design and validation of an asymmetric bowden-cable-driven series elastic actuator

Design and validation of an asymmetric bowden-cable-driven s...

引用

2019 IEEE International Conference on Robotics and Biomimetics, ROBIO 2019

作者： Sun, Ning Cheng, Long Tian, Lin Hou, Zeng-Guang Tan, Min Institute of Automation Chinese Academy of Sciences State Key Laboratory of Management and Control for Complex Systems Beijing100190 China University of Chinese Academy of Sciences China

ISBN: (纸本)9781728163215

Weight, size and torque control for hand exoskeleton robots are challenging due to lack of small and compact bidirectional torque actuators. In this paper, an asymmetric Bowden-cable-driven series elastic actuator (ABLE-SEA) for hand exoskeleton robots was proposed based on the fact that the required torques for finger movements are asymmetric. Compared to series elastic actuator and transmission system that were designed separately, the elastic elements of ABLESEA were placed in transmission parts to connect series elastic actuator with transmission system, which makes system compact. ABLE-SEA is 71 mm × 19.5mm × 20mm in dimension and weighs 30g except motor which is placed remotely. The dynamic model of ABLE-SEA was established, and feedback proportional-derivative (PD) control plus a feed-forward term was used to track the reference torque for the ABLE-SEA. The reference torque tracking test at different frequencies was performed at the developed prototype. Meanwhile, the peak torque of ABLE-SEA was tested. The experimental results verified that the torque bandwidth of the proposed series elastic actuator could reach 4Hz, and the peak torque of ABLE-SEA could reach 0.3Nm, which meet the design requirement. © 2019 IEEE.

关键词： Torque

来源：评论

学校读者我要写书评

暂无评论

Design of a robotic fish based on a passive flexible mechanism

Design of a robotic fish based on a passive flexible mechani...

引用

2019 IEEE International Conference on Robotics and Biomimetics, ROBIO 2019

作者： Lu, Ben Fu, Yuzhuo Zou, Qianqian Deng, Sai Zhou, Chao Institute of Automation Chinese Academy of Sciences State Key Laboratory of Management and Control for Complex Systems Beijing China University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781728163215

In recent years, along with the rapid development of manufacturing, control, sensing and other technologies, biomimetic robotic fish has played an increasingly important role in marine exploration, underwater rescue, water quality monitoring and other fields. This paper designed a robotic fish based on a passive flexible mechanism to realize the fish-like propulsion function. The motion simulation of the designed transmission mechanism was carried out to get the actual curve of the structure. Through the analysis of the force of the fishtail, the preliminary selection scheme of the elastic joints based on the passive flexible mechanism was obtained. Finally, this paper built a robotic fish based on a passive flexible mechanism, which could achieve the motion frequency adjustable from 0.5 to 9 Hz, and the maximum swimming speed reaching 1.05 BL/s (BL, body length). This design scheme provided a new reference for the design of the biomimetic robotic fish and had certain practical significance. © 2019 IEEE.

关键词： Fish

来源：评论

学校读者我要写书评

暂无评论

An Efficient Network for Lane Segmentation 4th

An Efficient Network for Lane Segmentation

引用

4th International Conference on Cognitive systems and Information Processing, ICCSIP 2018

作者： Li, Haoran Zhao, Dongbin Chen, Yaran Zhang, Qichao The State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9789811379826

As the basis of scenes understanding for autonomous driving, lane segmentation is always a challenge due to the various illumination conditions, heavy traffics and richly-textured roads. Because of the heavily biased distribution of lane/non-lane pixels, it is hard to achieve satisfying results by using image segmentation networks such as fully convolution neural networks (FCN). In this paper, we propose a new loss function to tackle the unbalanced data distribution problem. It has shown that the loss function significantly improves the performance of available segmentation networks such as FCN on the lane segmentation task. © 2019, Springer Nature Singapore Pte Ltd.

关键词： Autonomous vehicles

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：