Deep Learning Methods for Personnel Recognition based on Micro-Doppler Features

Shao, Yuming; Dai, Yulong; Yuan, Longzhi; Chen, Weidong

doi:10.1145/3163080.3163095

Cited by 11 publications

(5 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The above studies are also discussed in [ 28 ], where authors introduce the inception architecture to human gait micro-Doppler features for the first time. The obtained accuracy rate in persons recognition usind a CNN classifier is around 96.9%.…”

Section: Related Workmentioning

confidence: 99%

Temporal Convolutional Neural Networks for Radar Micro-Doppler Based Gait Recognition

Addabbo

Bernardi

Biondi

et al. 2021

Sensors

View full text Add to dashboard Cite

The capability of sensors to identify individuals in a specific scenario is a topic of high relevance for sensitive sectors such as public security. A traditional approach involves cameras; however, camera-based surveillance systems lack discretion and have high computational and storing requirements in order to perform human identification. Moreover, they are strongly influenced by external factors (e.g., light and weather). This paper proposes an approach based on a temporal convolutional deep neural networks classifier applied to radar micro-Doppler signatures in order to identify individuals. Both sensor and processing requirements ensure a low size weight and power profile, enabling large scale deployment of discrete human identification systems. The proposed approach is assessed on real data concerning 106 individuals. The results show good accuracy of the classifier (the best obtained accuracy is 0.89 with an F1-score of 0.885) and improved performance when compared to other standard approaches.

show abstract

Section: Related Workmentioning

confidence: 99%

Temporal Convolutional Neural Networks for Radar Micro-Doppler Based Gait Recognition

Addabbo

Bernardi

Biondi

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…Meanwhile, the spectrogram-based classifications of human aquatic activities and driver head motions were also investigated in [30] and [24] respectively, where date were generated from real measurements. Worth to mention, micro-Doppler signatures can also be used for personnel recognition as a personal identification [31]. Additionally, spectrograms were utilized for human gait and gesture classification in [32] and in [25] respectively.…”

Section: A Spectrogramsmentioning

confidence: 99%

Point Transformer-Based Human Activity Recognition Using High-Dimensional Radar Point Clouds

Guo

Guendel

Yarovoy

et al. 2023

2023 IEEE Radar Conference (RadarConf23)

View full text Add to dashboard Cite

My master's journey is coming to an end. When I look back on the two years of study, it has been a fantastic experience: pleasure with new friends, curiosity for new knowledge, and achievement upon finishing my master's project. Numerous things deeply impressed me. Despite all the restrictions due to the pandemic in the first year of study, I will always adore my study and life abroad in the Netherlands. Many thanks to everyone around me for making things better.First, I would like to express my gratitude to my supervisors: Dr. Francesco Fioranelli and Mr. Ronny Guendel. At the beginning of my thesis, thanks for providing a creative and comfortable atmosphere during weekly meetings, which allows me to explore the topic freely. When I got stuck, they always offered holistic and professional advice to help me push further. Also, they are enthusiastic about giving me solutions when my laptop goes down. The most impressive thing is the hands-on teaching of academic skills, such as guidance for my academic writing and suggestions for academic presentation.I am also grateful to Prof. Alexander Yarovoy. The first lesson I attended was the electromagnetics course taught by Prof. Yarovoy. His in-depth explanation of electromagnetic phenomena and cutting-edge technology attracted me deeply, which guided me to the MS3 group. I would also like to thank all the professors for their lectures and for expanding my horizon.My sincere thanks extend to Ph.D. candidates and MSc students in the MS3 group. I would like to show my gratitude to Simin, Sen, Yue, and Max for attending the midterm presentation rehearsal and giving me suggestions and encouragement. I am also grateful for deep and mindful discussions with Liyuan, Xingzhuo, Yue and Qinyu when I had questions about radars. In addition, I would like to thank Yubin, Yongdian, Xubin, and Ignacio for providing me with technical support. In addition, I would like to thank the secretary of MS3 for arranging my defense.Special thanks to my family. Without your support and effort, I would not have had the opportunity to study abroad, and your meticulous care fills my heart.The end of my master's journey is also the end of my student status. Still, my curiosity to explore nature and society will be like the fire of Prometheus, always shining.

show abstract

“…The second advantage allows pace or frequency information to be retained in the extracted features. In [23,[40][41][42][43][44][45][46][47][48][49], CNNs with different architectures and convolution kernels of various sizes were employed as classifiers to recognize human activities with time-Doppler maps, as illustrated in Figure 2a. In [28,[50][51][52], CNN acts as a spatial feature extractor and extracts high-level representations of human activities for further identification, as illustrated in Figure 2b.…”

Section: Convolutional Neural Networkmentioning

confidence: 99%

“…In addition, time-Doppler maps are intuitive and explicable. As a consequence, compared with other 2D radar echoes, the time-Doppler maps are most commonly used for radar-based HAR up to now [20,[41][42][43]45,48,49,54,56,69]. R.P.…”

Section: Deep Learning Approaches In 2d Radar Echomentioning

confidence: 99%

A Survey of Deep Learning-Based Human Activity Recognition in Radar

Jing

2019

Remote Sensing

234

View full text Add to dashboard Cite

Radar, as one of the sensors for human activity recognition (HAR), has unique characteristics such as privacy protection and contactless sensing. Radar-based HAR has been applied in many fields such as human–computer interaction, smart surveillance and health assessment. Conventional machine learning approaches rely on heuristic hand-crafted feature extraction, and their generalization capability is limited. Additionally, extracting features manually is time–consuming and inefficient. Deep learning acts as a hierarchical approach to learn high-level features automatically and has achieved superior performance for HAR. This paper surveys deep learning based HAR in radar from three aspects: deep learning techniques, radar systems, and deep learning for radar-based HAR. Especially, we elaborate deep learning approaches designed for activity recognition in radar according to the dimension of radar returns (i.e., 1D, 2D and 3D echoes). Due to the difference of echo forms, corresponding deep learning approaches are different to fully exploit motion information. Experimental results have demonstrated the feasibility of applying deep learning for radar-based HAR in 1D, 2D and 3D echoes. Finally, we address some current research considerations and future opportunities.

show abstract

Deep Learning Methods for Personnel Recognition based on Micro-Doppler Features

Cited by 11 publications

References 26 publications

Temporal Convolutional Neural Networks for Radar Micro-Doppler Based Gait Recognition

Temporal Convolutional Neural Networks for Radar Micro-Doppler Based Gait Recognition

Point Transformer-Based Human Activity Recognition Using High-Dimensional Radar Point Clouds

A Survey of Deep Learning-Based Human Activity Recognition in Radar

Contact Info

Product

Resources

About