Machine Learning (ML) and Deep Neural Network (DNN) based Computer-aided decision (CAD) systems show the effective implementation in solving skin cancer classification problem. However, ML approach unable to get the deep features from network flow which causes the low accuracy performance and the DNN model has the complex network with an enormous number of parameters that resulting in the limited classification accuracy. In this paper, the hybrid Convolutional Neural Network algorithm and Autoregressive Integrated Moving Average model (CNN-ARIMA) have been proposed to classify three different types of skin cancer. The proposed CNN-ARIMA able to classify skin cancer image successfully and achieved test accuracy, average sensitivity, average specificity, average precision and AUC of 96.00%, 96.02%, 97.98%, 96.13% and 0.995, respectively which outperformed the state-of-art methods.
{"title":"Skin Cancer Classification using Convolutional Neural Network with Autoregressive Integrated Moving Average","authors":"Chee Ka Chin, Dayang Azra binti Awang Mat, Abdulrazak Yahya Saleh","doi":"10.1145/3467691.3467693","DOIUrl":"https://doi.org/10.1145/3467691.3467693","url":null,"abstract":"Machine Learning (ML) and Deep Neural Network (DNN) based Computer-aided decision (CAD) systems show the effective implementation in solving skin cancer classification problem. However, ML approach unable to get the deep features from network flow which causes the low accuracy performance and the DNN model has the complex network with an enormous number of parameters that resulting in the limited classification accuracy. In this paper, the hybrid Convolutional Neural Network algorithm and Autoregressive Integrated Moving Average model (CNN-ARIMA) have been proposed to classify three different types of skin cancer. The proposed CNN-ARIMA able to classify skin cancer image successfully and achieved test accuracy, average sensitivity, average specificity, average precision and AUC of 96.00%, 96.02%, 97.98%, 96.13% and 0.995, respectively which outperformed the state-of-art methods.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134191618","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}
{"title":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","authors":"","doi":"10.1145/3467691","DOIUrl":"https://doi.org/10.1145/3467691","url":null,"abstract":"","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130708002","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}
In recent years, human parsing has been developed a lot for its valuable utilization. However, existing methods have not fully solved semantic errors and incomplete semantic predictions. In this regard, a Multi-Scale Feature Blend Network(MFBNet) is proposed to deal with these problems from the respective of fusing multi-scale features. Specifically, we creatively introduce the Context Embedding module which uses the feature pyramid as the main structure to blend multi-scale feature information. Besides, ResNet-101 is applied as the backbone network to train and optimize shared weights and map the generated feature maps to the Context Embedding module. Experimental results on several wide-used datasets show that the proposed method outperforms than the state-of-art methods in human parsing.
{"title":"A Novel Human Parsing Method Driven by Multi-Scale Feature Blend Network","authors":"Chunxu Wang, Benzhu Xu, Gaofeng Zhang","doi":"10.1145/3467691.3467692","DOIUrl":"https://doi.org/10.1145/3467691.3467692","url":null,"abstract":"In recent years, human parsing has been developed a lot for its valuable utilization. However, existing methods have not fully solved semantic errors and incomplete semantic predictions. In this regard, a Multi-Scale Feature Blend Network(MFBNet) is proposed to deal with these problems from the respective of fusing multi-scale features. Specifically, we creatively introduce the Context Embedding module which uses the feature pyramid as the main structure to blend multi-scale feature information. Besides, ResNet-101 is applied as the backbone network to train and optimize shared weights and map the generated feature maps to the Context Embedding module. Experimental results on several wide-used datasets show that the proposed method outperforms than the state-of-art methods in human parsing.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"72 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116657590","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}
Micromanipulation has been paid much attention due to its wide applications in micro/nano manufacturing and biological research in recent years. To achieve flexible micromanipulation in complex tasks, a macro-micro manipulator including 6-DOF macro-motion module and 3-DOF micro-motion module was presented. Based on the characteristics and demand of micromanipulation, a fast motion planning process for the macro-motion module was built. The collision detection model based on common sphere bounding box was improved. An example has been demonstrated to verify the collision detection model and motion planning process. The results showed that the proposed process could achieve a fast-planning speed and ensure smooth movement of the macro-motion module of the manipulator without collision.
{"title":"Motion planning of a macro-micro manipulator for flexible micromanipulation","authors":"Cheng Liu, Lefeng Wang, Jingyuan Liu, W. Rong","doi":"10.1145/3467691.3467700","DOIUrl":"https://doi.org/10.1145/3467691.3467700","url":null,"abstract":"Micromanipulation has been paid much attention due to its wide applications in micro/nano manufacturing and biological research in recent years. To achieve flexible micromanipulation in complex tasks, a macro-micro manipulator including 6-DOF macro-motion module and 3-DOF micro-motion module was presented. Based on the characteristics and demand of micromanipulation, a fast motion planning process for the macro-motion module was built. The collision detection model based on common sphere bounding box was improved. An example has been demonstrated to verify the collision detection model and motion planning process. The results showed that the proposed process could achieve a fast-planning speed and ensure smooth movement of the macro-motion module of the manipulator without collision.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"318 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126930921","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}
In this paper, a hierarchical-distributed control system of WLBOT is proposed. The control system is divided into three layers, the decision-making layer, the coordination layer, and the executive layer. The decision-making layer realizes human-computer interactive display and motion control instructions sending. The coordination layer realizes data coordination and movement planning. The execution layer realizes joint angle control and data collection. The communication between layers is based on WIFI and CAN bus. The control system uses the model-based control method to plan the movement trajectory and control WLBOT moving. Finally, the hierarchical-distributed control system is tested on WLBOT. From the experimental results, the control system can control the WLBOT moving correctly, and each foot is following the corresponding planning trajectory.
{"title":"The hierarchical-distributed control system of hydraulic walking robot WLBOT","authors":"Junkui Dong, Bo Jin, Ziqi Liu, Shuo Zhai","doi":"10.1145/3467691.3467702","DOIUrl":"https://doi.org/10.1145/3467691.3467702","url":null,"abstract":"In this paper, a hierarchical-distributed control system of WLBOT is proposed. The control system is divided into three layers, the decision-making layer, the coordination layer, and the executive layer. The decision-making layer realizes human-computer interactive display and motion control instructions sending. The coordination layer realizes data coordination and movement planning. The execution layer realizes joint angle control and data collection. The communication between layers is based on WIFI and CAN bus. The control system uses the model-based control method to plan the movement trajectory and control WLBOT moving. Finally, the hierarchical-distributed control system is tested on WLBOT. From the experimental results, the control system can control the WLBOT moving correctly, and each foot is following the corresponding planning trajectory.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"186 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124751909","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}
Currently, one of the diseases that is causing a large number of deaths in Peru is cervical cancer caused by the human papillomavirus (HPV). However, the application of the vaccine against this disease can protect against certain strains of HPV. The study consisted of the development of a predictive model using Machine Learning for the prognosis of HPV vaccination attrition in girls between 9 and 13 years of age. The data used came from the "HPV vaccination system" of the Peruvian Ministry of Health (MINSA). The methodology consisted of developing four supervised learning models: Decision Tree Classifier, Random Forest Classifier, Extra Trees Classifier and Extreme Gradient Boosting with the intention of comparing the results and choosing the best performing model for its respective calibration and to be used through a graphical interface. The results showed that the best learning model was Random Forest Classifier, with an Accuracy Score of 63.6140%, AUC of 63.6183%, Recall of 63% and F1-score of 63%; which indicates that the model classifies 64% of the cases as girls who drop out of the HPV vaccination program.
{"title":"Machine Learning-based predictive model for the prognosis of human papillomavirus (HPV) vaccination attrition","authors":"Urlish Marroquin, Nemias Saboya, A. Sullon","doi":"10.1145/3467691.3467695","DOIUrl":"https://doi.org/10.1145/3467691.3467695","url":null,"abstract":"Currently, one of the diseases that is causing a large number of deaths in Peru is cervical cancer caused by the human papillomavirus (HPV). However, the application of the vaccine against this disease can protect against certain strains of HPV. The study consisted of the development of a predictive model using Machine Learning for the prognosis of HPV vaccination attrition in girls between 9 and 13 years of age. The data used came from the \"HPV vaccination system\" of the Peruvian Ministry of Health (MINSA). The methodology consisted of developing four supervised learning models: Decision Tree Classifier, Random Forest Classifier, Extra Trees Classifier and Extreme Gradient Boosting with the intention of comparing the results and choosing the best performing model for its respective calibration and to be used through a graphical interface. The results showed that the best learning model was Random Forest Classifier, with an Accuracy Score of 63.6140%, AUC of 63.6183%, Recall of 63% and F1-score of 63%; which indicates that the model classifies 64% of the cases as girls who drop out of the HPV vaccination program.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134507341","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}
In order to realize the real-time detection of dairy goat objects in the sheep farm, a neural network detection algorithm based on depth wise separable convolution YOLOv3-DSN is proposed. Firstly, the video frames are used to screen out the key frames containing the dairy goats based on the surveillance video of the sheep farm, and construct the dairy goat sample set. Then the K-means clustering method is used to determine the number and dimensions of the object candidate box on the data set, and the GIOU box regression loss function is used to improve the positioning accuracy of the dairy goat regression box. At the same time, the model is optimized through multi-scale training, and the depth wise separable convolutionYOLOv3-DSN network is used to return the object category and position,which realizes end-to-end object detection.Under the circumstance of taking into account accuracy and speed, realize the object detection of sheep farm surveillance video.The experimental results show that compared with SSD and YOLOv3, it can obtain better object detection results in terms of efficiency and accuracy.Provide basic technology for the development of intelligent video surveillance systems for sheep farms and reduce the workload of experimenters.
{"title":"YOLOv3-DSN Object Detection Algorithm Based on Depth Wise Separable Convolution","authors":"Xujing Zhou, Jinglei Tang","doi":"10.1145/3467691.3467698","DOIUrl":"https://doi.org/10.1145/3467691.3467698","url":null,"abstract":"In order to realize the real-time detection of dairy goat objects in the sheep farm, a neural network detection algorithm based on depth wise separable convolution YOLOv3-DSN is proposed. Firstly, the video frames are used to screen out the key frames containing the dairy goats based on the surveillance video of the sheep farm, and construct the dairy goat sample set. Then the K-means clustering method is used to determine the number and dimensions of the object candidate box on the data set, and the GIOU box regression loss function is used to improve the positioning accuracy of the dairy goat regression box. At the same time, the model is optimized through multi-scale training, and the depth wise separable convolutionYOLOv3-DSN network is used to return the object category and position,which realizes end-to-end object detection.Under the circumstance of taking into account accuracy and speed, realize the object detection of sheep farm surveillance video.The experimental results show that compared with SSD and YOLOv3, it can obtain better object detection results in terms of efficiency and accuracy.Provide basic technology for the development of intelligent video surveillance systems for sheep farms and reduce the workload of experimenters.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127218779","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}
Jean Pierre Arce Misajel, Sario Angel Chamorro Quijano, Dominick Marco Cruz Esteban, Carlos Antonio Perea Fabian, Ruth Aracelis Manzanares Grados
This research presents the design and control of a mechatronic assistant for the treatment with music therapy in dementia patients, in emphasis on Alzheimer's disease, through control software implemented in a mechatronic system. The development of research shows that the proposed mechatronic system detects and records the behavior of brain waves, through communication in order to perform some corrective action against a possible unwanted unforeseen response during therapy. This communication can be in real time by communicating via Bluetooth through an application to a family member or patient manager and regulates the musical style according to the type of brain wave that best effects the patient with dementia and additionally, with a cloud record for further analysis of the improvement and progress of the patient. With this, the implemented mechatronic assistant will improve the well-being and quality of life of dementia patients.
{"title":"Design of a mechatronic assistant in the treatment of cognitive abilities using musical stimuli for people with dementia","authors":"Jean Pierre Arce Misajel, Sario Angel Chamorro Quijano, Dominick Marco Cruz Esteban, Carlos Antonio Perea Fabian, Ruth Aracelis Manzanares Grados","doi":"10.1145/3467691.3467705","DOIUrl":"https://doi.org/10.1145/3467691.3467705","url":null,"abstract":"This research presents the design and control of a mechatronic assistant for the treatment with music therapy in dementia patients, in emphasis on Alzheimer's disease, through control software implemented in a mechatronic system. The development of research shows that the proposed mechatronic system detects and records the behavior of brain waves, through communication in order to perform some corrective action against a possible unwanted unforeseen response during therapy. This communication can be in real time by communicating via Bluetooth through an application to a family member or patient manager and regulates the musical style according to the type of brain wave that best effects the patient with dementia and additionally, with a cloud record for further analysis of the improvement and progress of the patient. With this, the implemented mechatronic assistant will improve the well-being and quality of life of dementia patients.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"126 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126856317","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}
This paper focuses on the design and implementation of an anthropomorphic robotic finger, that was designed for teleoperation systems. This open-source anthropomorphic finger is easy to fabricate with a 3D printer and standard parts. The finger has three joints and two active Degrees of Freedom (dofs) with 2 servo motors dedicated to finger motion. Size and weight have been optimized in order to achieve human-like movement and space for attached feedback sensors. This proposed the design process, evaluate the finger with quantitative measures, both equation of four-bar linkage mechanism and equation of kinematic.
{"title":"Design of an Open Source Anthropomorphic Robotic Finger for Telepresence Robot","authors":"Jittaboon Trichada, Traithep Wimonrut, Narongsak Tirasuntarakul, Thanacha Choopojcharoen, Bawornsak Sakulkueakulsuk","doi":"10.1145/3467691.3467704","DOIUrl":"https://doi.org/10.1145/3467691.3467704","url":null,"abstract":"This paper focuses on the design and implementation of an anthropomorphic robotic finger, that was designed for teleoperation systems. This open-source anthropomorphic finger is easy to fabricate with a 3D printer and standard parts. The finger has three joints and two active Degrees of Freedom (dofs) with 2 servo motors dedicated to finger motion. Size and weight have been optimized in order to achieve human-like movement and space for attached feedback sensors. This proposed the design process, evaluate the finger with quantitative measures, both equation of four-bar linkage mechanism and equation of kinematic.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"78 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123175675","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：