∗There exist plenty of object detection applications in the field of remote sensing. However, some challenges appear especially for small and dense objects when the image is overlarge or the image with complex background. Thus, we propose a rotated attention wise network with an accuracy-speed balanced real-time objector. Learnable attentionwisemodules are adopted in the networkwith rotated bounding boxes for searching, locating right semantic and features simultaneously. In the process of training, various backbones and data augmentation strategies are employed to achieve higher accuracy and more types of objects. The results that conducted on extensive comparable experiments demonstrate the effectiveness of the proposed framework, achieving state-of-the-art performance in real-time object detection methods.
{"title":"Rotated YOLOv4 with Attention-wise Object Detectors in Aerial Images","authors":"Zhichao Zhang, Jinsheng Deng, Hui Chen, Xiaoqing Yin","doi":"10.1145/3467691.3467706","DOIUrl":"https://doi.org/10.1145/3467691.3467706","url":null,"abstract":"∗There exist plenty of object detection applications in the field of remote sensing. However, some challenges appear especially for small and dense objects when the image is overlarge or the image with complex background. Thus, we propose a rotated attention wise network with an accuracy-speed balanced real-time objector. Learnable attentionwisemodules are adopted in the networkwith rotated bounding boxes for searching, locating right semantic and features simultaneously. In the process of training, various backbones and data augmentation strategies are employed to achieve higher accuracy and more types of objects. The results that conducted on extensive comparable experiments demonstrate the effectiveness of the proposed framework, achieving state-of-the-art performance in real-time object detection methods.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"18 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":"133642098","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}
A new fast parallel constrained Viterbi algorithm for big data is proposed in this paper. We provide a detailed analysis of its performance on big data frameworks. This performance analysis includes the evaluation of execution time, speedup, and prediction accuracy. Additionally, we compare the impact of the proposed approach on the performance of our parallel constrained algorithm with other benchmark versions. We use synthetic data and real-world data in our experiments to describe the behavior of our algorithm for different data sizes and different numbers of nodes. We demonstrate that this algorithm is fast, highly efficient, and scalable when it runs on spark framework and its prediction quality is acceptable since there is no deterioration or reduction observed.
{"title":"Fast Parallel Constrained Viterbi Algorithm for Big Data with Applications to Financial Time Series","authors":"Imad Sassi, S. Anter, A. Bekkhoucha","doi":"10.1145/3467691.3467697","DOIUrl":"https://doi.org/10.1145/3467691.3467697","url":null,"abstract":"A new fast parallel constrained Viterbi algorithm for big data is proposed in this paper. We provide a detailed analysis of its performance on big data frameworks. This performance analysis includes the evaluation of execution time, speedup, and prediction accuracy. Additionally, we compare the impact of the proposed approach on the performance of our parallel constrained algorithm with other benchmark versions. We use synthetic data and real-world data in our experiments to describe the behavior of our algorithm for different data sizes and different numbers of nodes. We demonstrate that this algorithm is fast, highly efficient, and scalable when it runs on spark framework and its prediction quality is acceptable since there is no deterioration or reduction observed.","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":"132286144","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}
Nowadays, autonomous driving is becoming more and more popular. Lane line detection is very important for trajectory planning and decision making in autonomous driving. Traditional lane detection methods rely on highly defined, manual feature extraction and heuristic methods, which usually require post-processing technology. More and more recently, the approach is modeling with deep learning. The lane line scheme based on segmentation usually requires large model and complex convolution structure design, and it cannot perceive the lane line geometric features. Similar to the heat map scheme, the detection of the key points of the lane line actually belongs to the same scheme as the segmentation in a certain angle, but it only reduces part of the amount of computation. The current methods all ignore the data imbalance between the lane line categories that the near lane line occupies most of the position of the picture, resulting in far lane line samples are far less than the near samples. In this paper, a novel detection scheme for key points of lane lines is proposed. The key points of lane lines are linearly sampled at different intervals on the longitudinal axis of images to solve the problem of data imbalance between lane lines. Then the sampled anchor points are fixed, and the model only needs to predict the abscissa of each lane line at the anchor points. At the same time, the geometric constraint loss function of the lane line is put forward to ensure the correct lane line shape. The method presented in this paper achieves 50 FPS on embedded devices, it achieved SOTA on the Culane and Tusimple datasets.
{"title":"End-to-End Lane Detection: a Key Point Approach","authors":"Chuan Lv, Jinglei Tang, Ruoqi Wang","doi":"10.1145/3467691.3467696","DOIUrl":"https://doi.org/10.1145/3467691.3467696","url":null,"abstract":"Nowadays, autonomous driving is becoming more and more popular. Lane line detection is very important for trajectory planning and decision making in autonomous driving. Traditional lane detection methods rely on highly defined, manual feature extraction and heuristic methods, which usually require post-processing technology. More and more recently, the approach is modeling with deep learning. The lane line scheme based on segmentation usually requires large model and complex convolution structure design, and it cannot perceive the lane line geometric features. Similar to the heat map scheme, the detection of the key points of the lane line actually belongs to the same scheme as the segmentation in a certain angle, but it only reduces part of the amount of computation. The current methods all ignore the data imbalance between the lane line categories that the near lane line occupies most of the position of the picture, resulting in far lane line samples are far less than the near samples. In this paper, a novel detection scheme for key points of lane lines is proposed. The key points of lane lines are linearly sampled at different intervals on the longitudinal axis of images to solve the problem of data imbalance between lane lines. Then the sampled anchor points are fixed, and the model only needs to predict the abscissa of each lane line at the anchor points. At the same time, the geometric constraint loss function of the lane line is put forward to ensure the correct lane line shape. The method presented in this paper achieves 50 FPS on embedded devices, it achieved SOTA on the Culane and Tusimple datasets.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"48 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":"127190487","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}
Whether the air defense radar network layout is reasonable or not directly affects the combat effectiveness. The traditional layout method is mainly based on manual layout, which is greatly limited, so it is difficult to maximize the combat effectiveness of the whole system. In this paper, an improved particle swarm optimization algorithm is proposed to solve the problem of radar network layout optimization, which has the advantages of high speed and high precision. It can not only meet the needs of real-time simulation of electronic air defense operations, but also be used to assist commanders in decision-making in actual battlefield.
{"title":"Research on Intelligent station layout optimization of air defense radar network","authors":"Jun Li, Wen-Qi Dai, Lei Hu","doi":"10.1145/3467691.3467694","DOIUrl":"https://doi.org/10.1145/3467691.3467694","url":null,"abstract":"Whether the air defense radar network layout is reasonable or not directly affects the combat effectiveness. The traditional layout method is mainly based on manual layout, which is greatly limited, so it is difficult to maximize the combat effectiveness of the whole system. In this paper, an improved particle swarm optimization algorithm is proposed to solve the problem of radar network layout optimization, which has the advantages of high speed and high precision. It can not only meet the needs of real-time simulation of electronic air defense operations, but also be used to assist commanders in decision-making in actual battlefield.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"4 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":"128285914","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 presents an embedded controller for the quadruped hydraulic robot WLBOT. First, we give an overview of a WLBOT system. Second, the hardware design and the software architecture of the embedded controller are introduced. The embedded controller takes charge of multi-sensor information processing and signal output of the servo valve, as well as receiving control command and sending processed information via Control Area Network (CAN) bus. What's more, the realization of the 2kHz high-speed control of the embedded controller is illustrated. Finally, the platform is constructed, in which the feasibility of the design and the validity of the control algorithm is verified. It shows that WLBOT can walk properly in a PID controller as expected.
{"title":"An embedded controller for the hydraulic walking robot WLBOT","authors":"Ziqi Liu, Bo Jin, Shuo Zhai, Junkui Dong","doi":"10.1145/3467691.3467703","DOIUrl":"https://doi.org/10.1145/3467691.3467703","url":null,"abstract":"This paper presents an embedded controller for the quadruped hydraulic robot WLBOT. First, we give an overview of a WLBOT system. Second, the hardware design and the software architecture of the embedded controller are introduced. The embedded controller takes charge of multi-sensor information processing and signal output of the servo valve, as well as receiving control command and sending processed information via Control Area Network (CAN) bus. What's more, the realization of the 2kHz high-speed control of the embedded controller is illustrated. Finally, the platform is constructed, in which the feasibility of the design and the validity of the control algorithm is verified. It shows that WLBOT can walk properly in a PID controller as expected.","PeriodicalId":159222,"journal":{"name":"Proceedings of the 2021 4th International Conference on Robot Systems and Applications","volume":"94 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":"130512687","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：