HAND GESTURE RECOGNITION USING FMCW RADAR BASED ON CROSS-CONNECTION CONVOLUTIONAL NEURAL NETWORK
229 viewsDOI:
https://doi.org/10.54939/1859-1043.j.mst.75.2021.15-22Keywords:
Convolutional neural network; Hand gesture recognition; FMCW radar micro-Doppler spectrum.Abstract
This study proposes a cross-connection convolutional neural network, namely Cross‑CNN, to recognize hand gestures based on micro-Doppler spectrum data of FMCW (Frequency Modulated Continuous Wave) radar. In addition, different noise levels are added to the dataset for improving the recognition accuracy of the proposed model when predicting gestures in different noise conditions. The experimental results show that the model trained on the dataset with noise gives better recognition accuracy than the model trained on the dataset without noise does. Afterwards, the Cross-CNN model is investigated in changing the structural superparameters for selecting the most suitable parameter set for the proposed problem. Finally, the chosen Cross-CNN model is compared with other existing models in the same dataset and training conditions. As a result, the Cross-CNN network outperforms other models in terms of recognition accuracy, time-consumption and structural capacity thanks to using cross-connections which allow to combine new features with former ones in training process of the network.
References
[1]. X. Zabulis, H. Baltzakis, and A. A. Argyros, “Vision-based hand gesture recognition for human–computer interaction,” in The Universal Access Handbook. Boca Raton, FL, USA: CRC Press, 2009.
[2]. S. Ahmed, K. D. Kallu, S. Ahmed, and S. H. Cho, “Hand Gestures Recognition Using Radar Sensors for Human-Computer-Interaction: A Review,” Remote Sensing, vol. 13, no. 3, p. 527, Feb. 2021.
[3]. [Online] Available: https://atap.google.com/soli.
[4]. M. Scherer, M. Magno, J. Erb, P. Mayer, M. Eggimann and L. Benini, "TinyRadarNN: Combining Spatial and Temporal Convolutional Neural Networks for Embedded Gesture Recognition With Short Range Radars," in IEEE Internet of Things Journal, vol. 8, no. 13, pp. 10336-10346, 1 July1, 2021.
[5]. M. G. Amin, Z. Zeng and T. Shan, "Hand Gesture Recognition based on Radar Micro-Doppler Signature Envelopes," 2019 IEEE Radar Conference (RadarConf), 2019, pp. 1-6.
[6]. M. Ritchie, R. Capraru, and F. Fioranelli, “Dop-NET: a micro-Doppler radar data challenge,” Elec-tronics Letters, vol. 56, no. 11, pp. 568–570, May 2020.
[7]. Q. Zhang, “Micro-Doppler Characteristics of Radar Targets”, Kidlington, United Kingdom: Butterworth-Heinemann, 2016.
[8]. A. Zhang, Z. C. Lipton, M. Li, and A. J. Smola, “Dive into Deep Learning”. 2019.
[9]. K. Janocha and W. M. Czarnecki, “On Loss Functions for Deep Neural Networks in Classification,” Schedae Informaticae, vol. 1/2016, 2017.
[10]. S. Ruder, “An overview of gradient descent optimization algorithms,” arXiv:1609.04747v2, online [Available] https://arxiv.org/abs/1609.04747, 2016.
[11]. C. Szegedy et al., “Going Deeper with Convolutions,” arXiv:1409.4842v1 17 Sep 2014. [Online] Available: https://arxiv.org/pdf/1409.4842.pdf.
[12]. K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 770-778.
[13]. B. Zoph, V. Vasudevan, J. Shlens and Q. V. Le, “Learning Transferable Architectures for Scalable Image Recognition,” 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 8697-8710.
[14]. M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L.-C. Chen, “MobileNetV2: Inverted Residuals and Linear Bottlenecks,” IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 4510-4520.
