Improving UAV Personalized-Tracking Services by Fusing Visual and Radio Data

This work presents a high precision unmanned aerial vehicle (UAV) communications to detect, track and locate people during moving using real-time dataset. We use a fusion of visual and radio data collected using a UAV, where the UAV is used to provide radio signals to mobile users and to collect visual and wireless datasets simultaneously. During data collection, the UAV is connected to smartphones and laptops, and the UAV is connected to a controller and an edge server through a wireless network. You only look once version4 (YOLOv4), Kalman filter, and long-short term memory (LSTM) algorithms are combined to evaluate the proposed system. YOLOv4-Kalman filter uses video inputs to detect and track people. The detail information of the bounding boxes of the detected persons is integrated with wireless data to locate people in motion. We use a variety of optimization methods, such as batch normalization, error smoothing and DROPOUT to optimize the performances of our proposed system. The proposed model improves the conventional approach by 10 % during localization.

[1]  Ramakant Nevatia,et al.  Multiple Target Tracking by Learning-Based Hierarchical Association of Detection Responses , 2013, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Rui Zhang,et al.  Wireless communications with unmanned aerial vehicles: opportunities and challenges , 2016, IEEE Communications Magazine.

[3]  Sven Behnke,et al.  Online visual robot tracking and identification using deep LSTM networks , 2017, IEEE/RJS International Conference on Intelligent RObots and Systems.

[4]  Konrad Schindler,et al.  Online Multi-Target Tracking Using Recurrent Neural Networks , 2016, AAAI.

[5]  Chao Deng,et al.  Multiobject Tracking in Videos Based on LSTM and Deep Reinforcement Learning , 2018, Complex..

[6]  Chen Zhang,et al.  A hybrid outdoor localization scheme with high-position accuracy and low-power consumption , 2018, EURASIP J. Wirel. Commun. Netw..

[7]  Doo-Hyun Choi,et al.  Performance Enhancement of YOLOv3 by Adding Prediction Layers with Spatial Pyramid Pooling for Vehicle Detection , 2018, 2018 15th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS).

[8]  Dushyant Rao,et al.  Deep tracking in the wild: End-to-end tracking using recurrent neural networks , 2018, Int. J. Robotics Res..

[9]  Charles A. Bouman,et al.  Deep Learning for Moving Object Detection and Tracking from a Single Camera in Unmanned Aerial Vehicles (UAVs) , 2018, IMAWM.

[10]  Xiangyang Hao,et al.  Multi-target tracking of surveillance video with differential YOLO and DeepSort , 2019, International Conference on Digital Image Processing.

[11]  Hsin-Piao Lin,et al.  Mobility Prediction in Wireless Networks Using Deep Learning Algorithm , 2019, Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering.

[12]  David W. Matolak,et al.  A Survey of Air-to-Ground Propagation Channel Modeling for Unmanned Aerial Vehicles , 2018, IEEE Communications Surveys & Tutorials.

[13]  Wei Zhang,et al.  Coarse-to-Fine UAV Target Tracking With Deep Reinforcement Learning , 2019, IEEE Transactions on Automation Science and Engineering.

[14]  Dinesh Manocha,et al.  RobustTP: End-to-End Trajectory Prediction for Heterogeneous Road-Agents in Dense Traffic with Noisy Sensor Inputs , 2019, CSCS.

[15]  Ulus Çevik,et al.  Implementation of Target Tracking Methods on Images Taken from Unmanned Aerial Vehicles , 2019, 2019 IEEE 17th World Symposium on Applied Machine Intelligence and Informatics (SAMI).

[16]  Ryosuke Shibasaki,et al.  Pedestrian Trajectory Prediction in Extremely Crowded Scenarios , 2019, Sensors.

[17]  Fei Meng,et al.  A YOLOv3-based non-helmet-use detection for seafarer safety aboard merchant ships , 2019 .

[18]  Jingjing Wang,et al.  A Comprehensive Survey on UAV Communication Channel Modeling , 2019, IEEE Access.

[19]  Hong-Yuan Mark Liao,et al.  YOLOv4: Optimal Speed and Accuracy of Object Detection , 2020, ArXiv.

[20]  Miran Pobar,et al.  Tracking Handball Players with the DeepSORT Algorithm , 2020, ICPRAM.

[21]  Li-Chun Wang,et al.  Mobility Predictions for IoT Devices Using Gated Recurrent Unit Network , 2020, IEEE Internet of Things Journal.

[22]  Wei Liu,et al.  End-to-End Active Object Tracking and Its Real-World Deployment via Reinforcement Learning , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.