2018 IEEE International Conference on Robotics and Biomimetics (ROBIO)最新文献_第3页

An Automatic Scoliosis Diagnosis and Measurement System Based on Deep Learning 基于深度学习的脊柱侧凸自动诊断与测量系统

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8665296

Zhiqiang Tan, Kai Yang, Yu Sun, Bo Wu, Huiren Tao, Ying Hu, Jianwei Zhang

{"title":"An Automatic Scoliosis Diagnosis and Measurement System Based on Deep Learning","authors":"Zhiqiang Tan, Kai Yang, Yu Sun, Bo Wu, Huiren Tao, Ying Hu, Jianwei Zhang","doi":"10.1109/ROBIO.2018.8665296","DOIUrl":"https://doi.org/10.1109/ROBIO.2018.8665296","url":null,"abstract":"Adolescent idiopathic scoliosis (AIS) is a three-dimensional structural deformity of the spine which affects 1–4% of adolescents and causes not only deformed appearance but also compromised mental status, pulmonary function, motor function and life quality. Currently, the diagnosis of AIS depends on the measurement of Cobb angle on spine radiographs, which is performed manually by doctors. Intra-observer and inter-observer variation exist in such method and causes errors in diagnosis. The purpose of this study is to design an automatic scoliosis diagnosis and measurement system based on deep learning to improve measurement accuracy and assist doctors in diagnosis. U-net segmentation network was used to segment the spine radiographs. Based on the segmented images, specific positions of upper and lower end vertebrae (UEV and LEV) and slopes of their endplates were identified by minimum outer envelope rectangle and least square method. Subsequently, Cobb angles were measured as angles between superior endplates of UEVs and inferior endplates of LEVs. After comparing manually measured Cobb angles with computer- measured Cobb angles, the result showed that the computer- measured Cobb angles were similar to the manually measured ones. To sum up, this system for automatic measurement of Cobb angle can assist doctors in clinical diagnosis.","PeriodicalId":417415,"journal":{"name":"2018 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128581812","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: 7

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8665154

Petra Durovic, Marko Filipovic, R. Cupec

引用次数: 4

Collision Free Navigation of a Flying Robot for Underground Mine Search and Mapping 地下矿山搜索与测绘飞行机器人的无碰撞导航

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8665108

Hang Li, A. Savkin, B. Vucetic

引用次数: 3

Nondestructive Image De-Blurring Based on Diffraction Blurring Model 基于衍射模糊模型的无损图像去模糊

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8664752

Yangjie Wei, Jieqiong Du, Yongjun Liu

{"title":"Nondestructive Image De-Blurring Based on Diffraction Blurring Model","authors":"Yangjie Wei, Jieqiong Du, Yongjun Liu","doi":"10.1109/ROBIO.2018.8664752","DOIUrl":"https://doi.org/10.1109/ROBIO.2018.8664752","url":null,"abstract":"Image de-blurring is an important research branch in computer vision. Deconvolution methods, which deconvolute the blurred images with a degradation function (or point spread function) according to estimation of the blurring cause, are commonly used in the image de-blurring on macro-scale. However, these methods are difficult to deblur an image captured by a high-magnification microscopy, where the depth-of-field of the microscopy is limit and optical diffraction is obvious, because both depth variation and optical diffraction can result in blurring imaging. Due to the complicated coupling of depth and optical diffraction in micro/nano blurring imaging, the degradation function of each point may be different, and it is not reasonable to estimate it in the geometrical optics where optical diffraction is not considered. Therefore, the accuracy of these deconvolution methods is limit because their degradation functions do not include the influence of optical diffraction. In this paper, we researched the image blurring degradation process based on the theoretical relationship between the blurring degree and the depth variation, as well as optical diffraction, and then proposed an automatic method to calculate the degradation function of every pixel with a relationship between depth information and blurring degree. Finally, a non-destructive image de-blurring method was proposed and validated with different micro/nano scale samples. The experimental result proved the effectiveness and precision of our method.","PeriodicalId":417415,"journal":{"name":"2018 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124650036","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: 0

Bionic Visual-based Data Conversion for SLAM 基于仿生视觉的SLAM数据转换

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8665130

Mingzhu Li, Weimin Zhang, Yongliang Shi, Z. Yao, Zhenshuo Liang, Qiang Huang

{"title":"Bionic Visual-based Data Conversion for SLAM","authors":"Mingzhu Li, Weimin Zhang, Yongliang Shi, Z. Yao, Zhenshuo Liang, Qiang Huang","doi":"10.1109/ROBIO.2018.8665130","DOIUrl":"https://doi.org/10.1109/ROBIO.2018.8665130","url":null,"abstract":"Simultaneous localization and mapping (SLAM) is the key function for most mobile robots to achieve autonomous navigation. The traditional visual SLAM uses the camera to acquire data and constructs a sparse or dense 3D map, which is convenient for robot localization but difficult for obstacle avoidance and autonomous navigation. Thus, we propose an innovative data conversion algorithm based on bionic visual characteristics which can construct a two-dimensional accurate map for indoor navigation in this paper. The algorithm has two main parallel threads: Ground Detection and Data Conversion. The ground detection thread detects the ground in real time, and gets the transformation matrix from the camera to the ground based on the geometrical invariability. The data conversion thread first filters the depth data, and then proposes a variable-resolution model based on human visual characteristics, which can keep the conversion time consumption at a low level without affecting the accuracy. Each group of experiments shows that the data converted by our algorithm have high-precision, and can be used to construct the map for navigation accurately.","PeriodicalId":417415,"journal":{"name":"2018 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129659017","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: 0

Humanoid Robot Gait Planning Based on Virtual Supporting Point 基于虚拟支撑点的仿人机器人步态规划

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8664876

Xueheng Zhang, Mingguo Zhao

引用次数: 0

The characteristics of human-robot coadaptation during human-in-the-loop optimization of exoskeleton control 外骨骼控制人在环优化过程中人机协同适应特性研究

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8665057

Wen Wang, Yige Liu, Pengqing Ren, Juanjuan Zhang, Jingtai Liu

{"title":"The characteristics of human-robot coadaptation during human-in-the-loop optimization of exoskeleton control","authors":"Wen Wang, Yige Liu, Pengqing Ren, Juanjuan Zhang, Jingtai Liu","doi":"10.1109/ROBIO.2018.8665057","DOIUrl":"https://doi.org/10.1109/ROBIO.2018.8665057","url":null,"abstract":"Human-in-the-loop (HITL) optimization of exoskeleton control during assisted walking can improve human mobility and reduce the energy cost. This process involves human-robot coadaptation as suggested by prior studies. There was a drop in the same subjects metabolic cost under the same assisted walking condition before and after the optimization process. It means the subjects adapted to walking with the exoskeleton while the exoskeleton learned the optimal control parameters for the subjects. We analyzed the process of human bodies learning to walk with an ankle exoskeleton, aiming to quantify the characteristics of human-robot coadaptation during HITL optimization of exoskeleton control. Data of eleven participants from prior experiments were utilized in this study. We identified similar sample conditions for each participant and investigated the trend of metabolic cost along with the HITL exoskeleton control optimization process. Results showed that the relationship between human metabolic cost and the time past in the optimization cycle approximately followed exponential curves with widespread adaptation rates. For the optimization process of four parameters with each condition sampled for two minutes, the time constants were averaged at 238 ± 207 optimization sample conditions. Our results can provide guidance to the training process of robot assisted human motion.","PeriodicalId":417415,"journal":{"name":"2018 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"134 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129175430","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

Validation of a small flying e-nose system for air pollutants control: A plume detection case study from an agricultural machine 用于空气污染物控制的小型飞行电子鼻系统的验证:来自农业机械的羽流检测案例研究

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8664718

J. Valente, Sandra Munniks, Imke de Man, L. Kooistra

引用次数: 5

A Grasp Pose Detection Scheme with an End-to-End CNN Regression Approach 一种基于端到端CNN回归的抓取姿态检测方案

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8665219

Hu Cheng, M. Meng

{"title":"A Grasp Pose Detection Scheme with an End-to-End CNN Regression Approach","authors":"Hu Cheng, M. Meng","doi":"10.1109/ROBIO.2018.8665219","DOIUrl":"https://doi.org/10.1109/ROBIO.2018.8665219","url":null,"abstract":"In this paper, we proposed a solution to the problem of grasp pose detection with a convolutional neural network (CNN) trained and tested on the Cornell Grasp Dataset. We treat this task as a regression problem so that our network outputs the location, rotation and size of the grasp directly in a RGB or RGBD image. A novel loss is defined in the back propagation that makes the network select the grasp closest to the ground truth. This loss can prevent the predicted grasp from falling into the average location of the multiple grasp ground truth. We train the network by two cascade steps to make the network learn to predict the locations and rotations of the grasp, respectively. Because the prediction of the rotation is relatively difficult for the objects with irregular shapes, the weights for the loss of the grasp angle are increased during the second step by multiplying a scale factor. The proposed training process is simple and the pipeline is clean as our model is trained from end to end. We achieved a 90.4% grasp prediction accuracy in our experiments. In addition, we proposed a joint training network that generates quantity grasp candidates and classifies them as good or not good for the multiple grasp predictions.","PeriodicalId":417415,"journal":{"name":"2018 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"78 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130672414","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: 8

Sparse Reward Based Manipulator Motion Planning by Using High Speed Learning from Demonstrations 基于稀疏奖励的高速示范学习机械臂运动规划

2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) Pub Date : 2018-12-01 DOI: 10.1109/ROBIO.2018.8665328

Guoyu Zuo, Jiahao Lu, Tingting Pan

{"title":"Sparse Reward Based Manipulator Motion Planning by Using High Speed Learning from Demonstrations","authors":"Guoyu Zuo, Jiahao Lu, Tingting Pan","doi":"10.1109/ROBIO.2018.8665328","DOIUrl":"https://doi.org/10.1109/ROBIO.2018.8665328","url":null,"abstract":"This paper proposed a high speed learning from demonstrations (LfD) method for sparse reward based motion planning problem of manipulator by using hindsight experience replay (HER) mechanism and deep deterministic policy gradient (DDPG) method. First, a demonstrations replay buffer and an agent exploration replay buffer are created for storing experience data, and the hindsight experience replay mechanism is subsequently used to acquire the experience data from the two replay buffers. Then, the deep deterministic policy gradient method is used to learn the experience data and finally fulfil the manipulator motion planning tasks under the sparse reward. Last, experiments on the pushing and pick-and-place tasks were conducted in the robotics environment in the gym. Results show that the training speed is increased to at least 10 times as compared to the deep deterministic policy gradient method without demonstrations data. In addition, the proposed method can effectively utilize the sparse reward, and the agent can quickly complete the task even under the low success rate of demonstrations data.","PeriodicalId":417415,"journal":{"name":"2018 IEEE International Conference on Robotics and Biomimetics (ROBIO)","volume":"235 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123884784","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