2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP)最新文献_第3页

Humanoid action imitation learning via boosting sample DQN in virtual demonstrator environment 虚拟演示环境下基于增强样本DQN的人形动作模仿学习

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827324

Rong Zhou, Zhisheng Zhang, Kunyyu Peng, Yang Mi, Xiangsheng Huang

{"title":"Humanoid action imitation learning via boosting sample DQN in virtual demonstrator environment","authors":"Rong Zhou, Zhisheng Zhang, Kunyyu Peng, Yang Mi, Xiangsheng Huang","doi":"10.1109/M2VIP.2016.7827324","DOIUrl":"https://doi.org/10.1109/M2VIP.2016.7827324","url":null,"abstract":"With the growth of modern industrial automation, autonomous-learning applied in the field of robot has aroused considerable attentions of researchers. However, those existing learning methods typically require mass among of training set, increasing the difficulty of collecting samples which is time-consuming, while the validity of samples might be divergent greatly, and thus the training efficiency is limited. Simultaneously, the reinforcement learning used in the system was based on the hypothesis that each action in the sequence contribute equally to the consequence, which is not corresponding to the common rules. In this paper, we propose a method, boosting sample DQN, to optimize the validity of training sample set. Inspired by boosting method, by extracting samples from replay memory hierarchically based on statistical results, the efficiency of network training is improved. Our algorithm, which has a small count of parameters, has been transplanted to the dual-arm robot system successfully. This approach learns a set of trajectories for the action of reaching and grabbing target objects using real-time models obtained by interactively wearable sensing equipment. And also, solution was proposed to distinguish weights of different actions. Our method has proved to be adaptive in learning complicated tasks, including grabbing bottle within its scope, as we presented in the paper.","PeriodicalId":125468,"journal":{"name":"2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125766893","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Mobile machine vision development for urine patch detection 尿斑检测的移动机器视觉开发

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827315

Akshaya Kumar, Hamid Sharifi, K. Arif

引用次数: 2

Two opposite sides synchronous tracking X-ray based robotic system for welding inspection 基于x射线的两对边同步跟踪焊接检测机器人系统

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827334

Kai Zheng, Jie Li, Chun Lei Tu, Xing Song Wang

引用次数: 6

Dynamic analysis of a cable-climbing robot system 攀索机器人系统的动力学分析

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827285

Wang Yue, Xu Fengyu, Yang Zhong

{"title":"Dynamic analysis of a cable-climbing robot system","authors":"Wang Yue, Xu Fengyu, Yang Zhong","doi":"10.1109/M2VIP.2016.7827285","DOIUrl":"https://doi.org/10.1109/M2VIP.2016.7827285","url":null,"abstract":"Cable-climbing robot is designed to inspect the cable-stayed bridge, releasing workers from dangerous environment. Meanwhile, the vibration of cable with a robot should be researched. The cables of bridge will be seriously damaged when large amplitude occurs. Not only that, the inertial force by acceleration should be emphasized as well. In this paper, a cable-climbing robot is introduced first. Then we analyze the vibration system of cable with this robot, regarding the robot as a particle and considering wind effects. We mainly consider two parameters, mass and position, which may strongly influence the vibration. Also, we refer to the Vibration String Equation, regarding the system as pure forced vibration with lumped mass. We give the simulation of the cable without robot for comparison. The result shows that maximum amplitude occurs in the middle of the cable. When the robot climbs on the cable, vibration by the lumped mass is discussed. Here we use a method of subsection, supplementing new condition at robot position. After solving the equations on both sides, we get a piecewise function that concludes mass and position which are important in further research.","PeriodicalId":125468,"journal":{"name":"2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP)","volume":"97 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114907281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Entity tracking within a Zigbee based smart home 基于Zigbee的智能家居中的实体跟踪

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827294

Daniel Konings, A. Budel, F. Alam, Frazer K. Noble

{"title":"Entity tracking within a Zigbee based smart home","authors":"Daniel Konings, A. Budel, F. Alam, Frazer K. Noble","doi":"10.1109/M2VIP.2016.7827294","DOIUrl":"https://doi.org/10.1109/M2VIP.2016.7827294","url":null,"abstract":"Modern Smart Home Automation (SMA) systems are predominantly based on wireless communication standards. New home automation setups can often include dozens of devices ranging from thermostats, humidity sensors, light switches, digital locks and cooling systems, all communicating on a common network. In this paper we focus on how SMA systems built on one of the most common standards (Zigbee), could be leveraged to provide a secondary benefit in the form of an Indoor Positioning System (IPS). IPS can be implemented in the form of Device Free Localization (DfL), Active tracking or as a combination of both techniques. A system containing a DfL implementation can detect and track moving entities by monitoring the changes in received signal strength (RSSI) values between nodes within a wireless network. DfL does not require the entity that is being tracked to carry an electronic device and actively contribute to the localization process. In Active tracking, the tracked entity contributes to the tracking process. In this paper both techniques are implemented individually, and a combination of both techniques is explored. Having implemented DfL and Active tracking, we were able to localize a person within a 3m × 3m quadrant with DfL with 80% accuracy. Active tracking resulted in a higher resolution of tracking compared to DfL, being able to localize a person within a 2m × 2m area with 95% accuracy. The accuracy of Active Tracking was then further increased to 98%, by coupling Active Tracking with DfL measurements.","PeriodicalId":125468,"journal":{"name":"2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133691399","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 10

Vision-based autonomous micro-air-vehicle control for odor source localization 基于视觉的气味源定位自主微型飞行器控制

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827276

K. Kurotsuchi, M. Tai, H. Takahashi

{"title":"Vision-based autonomous micro-air-vehicle control for odor source localization","authors":"K. Kurotsuchi, M. Tai, H. Takahashi","doi":"10.1109/M2VIP.2016.7827276","DOIUrl":"https://doi.org/10.1109/M2VIP.2016.7827276","url":null,"abstract":"This paper presents a novel control method for autonomous-odor-source localization using vision and odor sensing by micro air vehicles (MAVs). Our method is based on biomimetics, which enable highly autonomous localization. Our method does not need any instruction signals, including even global positioning system (GPS) signals. An experimenter just blows a whistle, and the MAV starts to hover, to seek an odor source, and to keep hovering near the source. The GPS-signal-free control based on vision enables indoor/underground use. Moreover, the MAV is light-weight (85 grams) and does not cause harm to others even if it accidentally falls. Experiments conducted in the real world were successful in enabling odor source localization using the MAV with a bio-inspired searching method. The distance error of the localization was 63 cm, more accurate than the target distance of 120 cm for individual identification. Our odor source localization is the first step to a proof of concept for a danger warning system. These results will be applied to the system to enable a safer and more secure society.","PeriodicalId":125468,"journal":{"name":"2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130361111","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 4

A novel double-layer, multi-channel soft pneumatic actuator that can achieve multiple motions 一种新型双层多通道气动软执行器，可实现多种运动

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827343

Qi Zhang, Zhisheng Zhang

引用次数: 4

Design of end-effectors for a chemistry automation plant 化工自动化装置末端执行器设计

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827304

Akshaya Kumar, Kamila Pillearachichige, Hamid Sharifi, Ben Shaw, Frazer K. Noble

引用次数: 2

Physical animation simulation of robot kit 机器人套件的物理动画仿真

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827313

Haitao Gao, Lei Zhang, Wenzheng Ding, Fei Hao, Yinglu Zhou

引用次数: 0

Vehicle detection in high resolution satellite images with joint-layer deep convolutional neural networks 基于联合层深度卷积神经网络的高分辨率卫星图像车辆检测

2016 23rd International Conference on Mechatronics and Machine Vision in Practice (M2VIP) Pub Date : 2016-11-01 DOI: 10.1109/M2VIP.2016.7827266

Yanjun Liu, Na Liu, H. Huo, T. Fang

引用次数: 3