Papers - Ohtsuki, Tomoaki

Non-contact Heartbeat Detection Based on Beam Diversity Using Multibeam Doppler Sensor

Kitagawa T., Yamamoto K., Ohtsuki T.

2021 IEEE Global Communications Conference, GLOBECOM 2021 - Proceedings (2021 IEEE Global Communications Conference, GLOBECOM 2021 - Proceedings)   2021

　View Summary

Heartbeat detection is a promising technology for health care and is in great demand in many applications. For non-contact and non-invasive heartbeat detection, Doppler sensor-based heartbeat detection methods have been investigated extensively. However, since the conventional methods use a single beam Doppler sensor, the accuracy of heartbeat detection tends to degrade, when the SNR (Signal-to-Noise Ratio) of heartbeat components for the beam direction is low. In this paper, we propose a non-contact heartbeat detection method based on beam diversity using a multibeam Doppler sensor. The effects of noises and the corresponding SNRs of heartbeat components over the received signal differ from one beam to another. Based on this fact, the proposed method detects heartbeat by exploiting beam diversity of the received signals. The experimental results showed that compared to the detection method using a single beam, our proposed method using multiple beams detected heartbeat accurately, which indicates that exploiting the beam diversity can lead to the improvement of heartbeat detection accuracy.

• Access to Document (DOI)

Multi-Rate Compression for Downlink CSI Based on Transfer Learning in FDD Massive MIMO Systems

Wang Y., Sun J., Wang J., Yang J., Ohtsuki T., Adebisi B., Gacanin H.

IEEE Vehicular Technology Conference (IEEE Vehicular Technology Conference)  2021-September 2021

ISSN  15502252

　View Summary

Accurate downlink channel state information (CSI) is one of the essential requirements for harnessing the potential advantages of frequency-division duplexing (FDD) massive multi-input multi-output (MIMO) systems. The current state-of-art in this vibrant research area include the use of deep learning to compress and feedback downlink CSI at the user equipments (UEs). These approaches focus mainly on achieving CSI feedback with high reconstruction performance and low complexity, but at the expense of inflexible compression rate (CR). High training overheads and limited storage capacity requirements are some of the challenges associated with the design of dynamic CR, which instantaneously adapt to propagation environment. This paper applies transfer learning (TL) to develop a multi-rate CSI compression and recovery neural network (TL-MRNet) with reduced training overheads. Simulation results are presented to validate the superiority of the proposed TL-MRNet over traditional methods in terms of normalized mean square error and cosine similarity.

• Access to Document (DOI)

Lightweight Network Design Based on ResNet Structure for Modulation Recognition

Lu X., Tao M., Fu X., Gui G., Ohtsuki T., Sari H.

IEEE Vehicular Technology Conference (IEEE Vehicular Technology Conference)  2021-September 2021

ISSN  15502252

　View Summary

The problem of unknown modulation signal recognition has been received intensely attentions in next-generational intelligent wireless communications. The deep learning (DL) has been widely used in unknown modulation signal recognition due to its excellent performance in solving classification problems and the DL-based automatic modulation classification (AMC) had been proposed. However, DL-based AMC method usually has high space complexity and computational complexity, which limits DL-based AMC to miniaturized devices with limited storage and computing capability. Therefore, a lightweight residual neural network (LResNet) for AMC is proposed in this paper. The simulation results show that the model parameters of LResNet is about 4.8% of the traditional CNN network, and about 14.9% of the ResNet and the classification performance of LResNet improves more than 3% compared with the traditional CNN network and decreases less than 1.5% compared to the ResNet.

• Access to Document (DOI)

Lightweight network and model aggregation for automatic modulation classification in wireless communications

Fu X., Gui G., Wang Y., Ohtsuki T., Adebisi B., Gacanin H., Adachi F.

IEEE Wireless Communications and Networking Conference, WCNC (IEEE Wireless Communications and Networking Conference, WCNC)  2021-March 2021

ISSN  15253511

　View Summary

This paper proposes a decentralized automatic modulation classification (DecentAMC) method using light network and model aggregation. Specifically, the lightweight network is designed by separable convolution neural network (S-CNN), in which the separable convolution layer is utilized to replace the standard convolution layer and most of the fully connected layers are cut off, the model aggregation is realized by a central device (CD) for edge device (ED) model weights aggregation and multiple EDs for ED model training. Simulation results show that the model complexity of S-CNN is decreased by about 94% while the average CCP is degraded by less than 1% when compared with CNN and that the proposed AMC method improves the training efficiency when compared with the centralized AMC (CentAMC) using S-CNN.

• Access to Document (DOI)

HSRRS Classification Method Based on Deep Transfer Learning and Multi-Feature Fusion

Wang Z., Li Z., Wang Y., Li W., Yang J., Ohtsuki T.

IEEE Vehicular Technology Conference (IEEE Vehicular Technology Conference)  2021-September 2021

ISSN  15502252

　View Summary

Convolutional neural network (CNN) is one of the most important tools to accomplish high-spatial-resolution remote sensing (HSRRS) image classification tasks with their unique feature extraction and feature expression capabilities. However, the CNN-based classification method is very limited due to the acquisition of HSRRS images is difficult and the sample size is limited. In addition, the extraction of features by a single model is very limited, which limits the further improvement of classification performance. To solve the above problems, we propose ResNet50-InceptionV3 based on deep transfer learning and multi-feature fusion (TLMFFRI) model to apply for high-spatial-resolution remote sensing image classification. First, both ResNet50 and InceptionV3 are trained on the ImageNet dataset. Then, transfer the trained convolutional layers weights to the TLMFFRI model to fuse the features and realize the HSRRS image classification. Finally, we evaluate the method on the HSRRS dataset. Compared with ResNet50 based on transfer learning (TL-ResNet50) and InceptionV3 based on transfer learning (TL-InceptionV3), the proposed method achieved better classification performance.

• Access to Document (DOI)

Fetal Heart Rate Detection Using First Derivative of ECG Waveform and Multiple Weighting Functions

Niida N., Wang L., Ohtsuki T., Owada K., Honma N., Hayashi H.

Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS (Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS)     434 - 438 2021

ISSN  1557170X

　View Summary

Fetal heart rate monitoring using the abdominal electrocardiograph (ECG) is an important topic for the diagnosis of heart defects. Many studies on fetal heart rate detection have been presented, however, their accuracy is still unsatisfactory. That is because the fetal ECG waveform is contaminated by maternal ECG interference, muscle contractions, and motion artifacts. One of the conventional methods is to detect the R-peaks from the integrated power of the frequency corresponding to the fetal heartbeats. However, the detection accuracy of the R-peaks is not enough. In this paper, we propose a method to generate the candidates of R-peaks using the first derivative of the signal and to pick up the estimated heartbeats by a multiple weighting function. The proposed multiple weighting function is designed by the Gaussian distribution, of which parameters are set from a grid search with the goal of minimizing the standard deviation of RR intervals (neighboring R-peaks intervals). The validation for the proposed framework has been evaluated on real-world data, which got the better accuracy than the conventional method that detects R-peaks from the integrated power and uses the weighting function produced by a fixed parameter of Gaussian distribution [12]. The averaged absolute error (AAE) which compares the estimated fetal heart rate and the reference fetal heart rate has been decreased by 17.528 bpm.

• Access to Document (DOI)

Fast Beamforming Design Method for IRS-Aided mmWave MISO Systems

He Z., Huang H., Yang J., Gui G., Ohtsuki T., Adebisi B., Gacanin H.

IEEE Vehicular Technology Conference (IEEE Vehicular Technology Conference)  2021-September 2021

ISSN  15502252

　View Summary

Intelligent reflecting surface (IRS)-aided millimeter-wave (mmWave) multiple-input single-output (MISO) is considered one of the promising techniques in next-generation wireless communication. However, existing beamforming methods for IRS-aided mm Wave MISO systems require high computational power, so it cannot be widely used. In this paper, we combine an unsupervised learning-based fast beamforming method with IRS-aided MISO systems, to significantly reduce the computational complexity of this system. Specifically, a new beamforming design method is proposed by adopting the feature fusion means in unsupervised learning. By designing a specific loss function, the beamforming can be obtained to make the spectrum more efficient, and the complexity is lower than that of the existing algorithms. Simulation results show that the proposed beamforming method can effectively reduce the computational complexity while obtaining relatively good performance results.

• Access to Document (DOI)

Discrete BP Polar Decoder Using Information Bottleneck Method

Yamada A., Ohtsuki T.

IEEE Access (IEEE Access)  9   10645 - 10656 2021

　View Summary

Polar code is one of the channel codes and is used in the 5th generation of mobile communication system (5G). This encoding scheme is based on the operation of channel polarization, and it is possible to achieve the capacity of arbitrary binary-input symmetric discrete memoryless channels. Compared with Turbo codes and LDPC (Low-Density Parity-Check) codes, the implementation of the encoder and decoder is easier. BP (Belief Propagation) decoding, one of the decoding methods of the polar codes, can be performed at high speed because it can decode in parallel. However, the disadvantages of the BP decoding are the hardware and computational complexities. The technique of quantization can be used to reduce complexities of hardware and calculation. One of the quantization methods is the information bottleneck method, which allows an observation variable to compressed one while trying to preserve the mutual information shared with a relevant variable. As a novel approach, the information bottleneck method is used in the design of quantizers for the BP decoding of LDPC codes. In this paper, we propose a discrete BP polar decoder that can use only unsigned integers in the decoding process by using the information bottleneck method. Thus, we can replace complex calculations of BP decoding with simple lookup tables. We also investigate the minimum bit width for quantization with negligible degradation and the suboptimal Eb/N0 for designing lookup tables, where Eb and N0 denote energy per bit and noise power density, respectively. The simulation results show that the proposed method can achieve almost the same error correcting capability compared with the BP decoding without compression in the range of low Eb/N0. Besides, we show that the proposed decoder can compress both channel outputs and BP messages with small loss compared with the uniform quantization decoder.

• Access to Document (DOI)

Differentiable architecture search-based automatic modulation classification

Wei X., Luo W., Zhang X., Yang J., Gui G., Ohtsuki T.

IEEE Wireless Communications and Networking Conference, WCNC (IEEE Wireless Communications and Networking Conference, WCNC)  2021-March 2021

ISSN  15253511

　View Summary

Automatic modulation classification (AMC) is an essential and meaningful technology in the development of cognitive radio. It can judge the modulation mode according to the signal acquired by the receiver. In recent years, the deep learning (DL) method has been used to take the place of modulation signal recognition based on decision theory and pattern recognition, which has achieved very effective results. The development of the neural network classification model focuses on architectural engineering. Discovering state-of-the-art neural network architectures requires substantial prior knowledge and effort of human experts. Neural architecture search (NAS) can be viewed as a subdomain of automatic machine learning (AutoML), which uses a neural network to automatically adjust the structures and parameters to obtain a network that researchers need by following search strategies that maximize performance. In this paper, we propose a differentiable architecture search (DARTS) based AMC method. In addition, we also consider six other methods, including convolutional neural network (CNN), simple recurrent unit (SRU), a convolutional-recurrent neural network (CRFN-CSS), Residual Networks (ResNet), Inception Modules (Inception) and MobileNet. Simulation results show that the proposed method can achieve the optimal classification accuracy at low parameters and floating-point operations (FLOPs) without manual architecture engineering.

• Access to Document (DOI)

Deep Learning-Based Channel Estimation for Massive MIMO Systems with Pilot Contamination

Hirose H., Ohtsuki T., Gui G.

IEEE Open Journal of Vehicular Technology (IEEE Open Journal of Vehicular Technology)  2   67 - 77 2021

　View Summary

In a time division duplex (TDD) based massive multiple-input multiple-output (MIMO) system, a base station (BS) is required to obtain accurate estimation of channel state information (CSI) for a user terminal (UT). Because of the time-varying nature of the channel, the length of pilot signals is limited and the number of orthogonal pilot signals is finite. Hence, the same pilot signals are required to be reused in neighboring cells and thus its channel estimation performance is deteriorated by pilot contamination from the neighboring cells. The minimum mean square error (MMSE) channel estimation can be used to reduce the influence of pilot contamination. However, it needs to know the covariance matrix of channels for all the UTs, which is unknown to the BS in practice. In this paper, we propose two methods of deep learning aided channel estimation to reduce the influence of pilot contamination. One method uses a neural network consisting of fully connected layers, while the other method uses a convolutional neural network (CNN). The neural network, particularly the CNN, plays a role in extracting features of the spatial information from the contaminated signals. The former method is better in terms of the training speed, however, the latter one can estimate the channel more accurately. We evaluate the proposed methods under two scenarios, i.e., perfect timing synchronization and imperfect one. Simulation results confirm that the proposed methods are better than the LS and covariance estimation methods via normalized mean square error (NMSE). In addition, we also investigate the impact of channel aging, and show that including some expected data into training datasets can avoid the great degradation of estimation quality.

• Access to Document (DOI)

Decentralized Learning-based Scenario Identification Method for Intelligent Vehicular Communications

Zhou Y., Wangt Y., Liu P., Yang J., Ohtsuki T., Sari H.

IEEE Vehicular Technology Conference (IEEE Vehicular Technology Conference)  2021-September 2021

ISSN  15502252

　View Summary

Scenario identification (SCI) is one of key techniques for intelligent vehicular communications (IVC) to maintain an effective and reliable operating state. Based on the deep learning (DL), it is a hotspot to identify scenarios of wireless communication using the characteristic quantity inherent in wireless channels. This paper proposes a decentralized learning-based SCI (DecentSCI) for IVC, relying on the algorithm of lightweight and model aggregation. By improving training efficiency and meanwhile reducing model complexity, the proposed method achieves low computing and communication, which is applicable for vehicular devices. Simulation results show that the training efficiency is upgraded by 97.15% and the model complexity is decreased by 90.25% at the cost of slight performance loss, i.e., 0.15%.

• Access to Document (DOI)

Surveillance Plane Aided Air-Ground Integrated Vehicular Networks: Architectures, Applications, and Potential

Sun J., Liu F., Zhou Y., Gui G., Ohtsuki T., Guo S., Adachi F.

IEEE Wireless Communications (IEEE Wireless Communications)  27 ( 6 ) 122 - 128 2020.12

ISSN  15361284

　View Summary

© 2002-2012 IEEE. Future air-ground integrated vehicular network (AGVN) is a vital part of the expected ubiquitous communication system. The resources in AGVN are geographically separated and are often heterogeneous. Therefore, distributed management and cooperative management are key issues in a multi-tier based topology of AGVN. Recently, artificial intelligence (AI) shows advantages in facilitating the management of AGVN; however, adequate, real-time, and secure information for driving AI is still a big concern. In this article, we propose a conceptual enhanced AGVN by introducing a so-called surveillance plane to promote the management in the control plane. Thus, components such as virtual network functions (VNFs) and service function chaining (SFC) can get profits. Concepts and trials of utilizing the surveillance plane and side information are presented from three different aspects, namely networking, security, and application. We also summarize research directions in designing a more flexible management scheme by coupling the service, surveillance, and control planes of the enhanced AGVN.

• Access to Document (DOI)

Dual-Ascent Inspired Transmit Precoding for Evolving Multiple-Access Spatial Modulation

Cao Y., Ohtsuki T., Quek T.Q.S.

IEEE Transactions on Communications (IEEE Transactions on Communications)  68 ( 11 ) 6945 - 6961 2020.11

ISSN  00906778

　View Summary

© 1972-2012 IEEE. In this article, we investigate the dual-ascent inspired transmit precoding (TPC) for multiple-access spatial modulation (MASM) in multiple-input multiple-output (MIMO) systems. Note that several novel TPC techniques have been developed in earlier works, to provide a solution for either the maximal-minimum Euclidean distance, or the quadratically constrained quadratic program problems. However, numerical results expose that the system performances degrade distinctly when applying these TPC techniques into MASM-MIMO. The main reason behind is that these TPC techniques are sensitive to the system dimensions and the quadratic constraints. In this context, we first recast the above challenging problems as an unconstrained problem by imposing a penalty over the quadratic constraints. Based on the primal-dual optimality theory, we next propose a Broyden-Fletcher-Goldfarb-Shanno (BFGS) aided dual-ascent approach for finding a global optimum solution to the unconstrained problem. Further, we introduce non-stationary time-varying TPC parameters to characterize an evolving MASM-MIMO in which the signals are multiplexed over a small coherence time, and thereby resulting in dual-ascent aided non-stationary TPC approach. Numerical results manifest that the proposed algorithms possess an inherent robustness to the increasing system dimension and quadratic constraint. Besides, simulation results show the benefits of our algorithms under different kinds of performance metrics.

• Access to Document (DOI)

Evaluating smart grid renewable energy accommodation capability with uncertain generation using deep reinforcement learning

Liu Y., Guan X., Li J., Sun D., Ohtsuki T., Hassan M.M., Alelaiwi A.

Future Generation Computer Systems (Future Generation Computer Systems)  110   647 - 657 2020.09

ISSN  0167739X

　View Summary

© 2019 Elsevier B.V. Due to environment-friendliness, renewable energy like solar power and wind power is more and more introduced to energy systems all over the world. Simultaneously, high penetrations of wind and solar generation also have brought severe curtailment of wind and solar. How to alleviate curtailment of wind and solar is a crucial problem in evaluating accommodation capability of renewable energy, which reflects the extent of utilization of renewable energy and economic benefits. The uncertainty of renewable energy brings challenges to precisely describe renewable generation, which leads to difficulty in designing effective mechanisms for accommodation capability of renewable energy. Existing work suffers from high computation overhead from frequently updated data, and low precision of describing renewable energy, which leads to less effective policies for renewable energy accommodation and underestimated accommodation capability. To make the most of renewable energy, an algorithm AccCap-DRL based on deep reinforcement learning is proposed. AccCap-DRL partitions a distribution into segments by time intervals, employs WGAN to describe distributions of renewable energy data, and employs DDPG to obtain approximate policies for renewable energy accommodation in different scenarios. Simulation results from real power generation and users’ demand data show high effectiveness of the proposed algorithm, and high efficiency of evaluating accommodation capability.

• Access to Document (DOI)

Semi-Supervised Machine Learning Aided Anomaly Detection Method in Cellular Networks

Lu Y., Wang J., Liu M., Zhang K., Gui G., Ohtsuki T., Adachi F.

IEEE Transactions on Vehicular Technology (IEEE Transactions on Vehicular Technology)  69 ( 8 ) 8459 - 8467 2020.08

ISSN  00189545

　View Summary

© 1967-2012 IEEE. The ever-increasing amount of data in cellular networks poses challenges for network operators to monitor the quality of experience (QoE). Traditional key quality indicators (KQIs)-based hard decision methods are difficult to undertake the task of QoE anomaly detection in the case of big data. To solve this problem, in this paper, we propose a KQIs-based QoE anomaly detection framework using semi-supervised machine learning algorithm, i.e., iterative positive sample aided one-class support vector machine (IPS-OCSVM). There are four steps for realizing the proposed method while the key step is combining machine learning with the network operator's expert knowledge using OCSVM. Our proposed IPS-OCSVM framework realizes QoE anomaly detection through soft decision and can easily fine-Tune the anomaly detection ability on demand. Moreover, we prove that the fluctuation of KQIs thresholds based on expert knowledge has a limited impact on the result of anomaly detection. Finally, experiment results are given to confirm the proposed IPS-OCSVM framework for QoE anomaly detection in cellular networks.

• Access to Document (DOI)

An Infrared Array Sensor-Based Method for Localizing and Counting People for Health Care and Monitoring

Bouazizi M., Ohtsuki T.

Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS (Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS)  2020-July   4151 - 4155 2020.07

ISSN  9781728119908

　View Summary

© 2020 IEEE. To build a system for monitoring elderly people living alone, an important step needs to be done: identifying the presence/absence of the person being monitored and his location. Such task has several applications that we discuss in this paper, and remains very important. Several techniques were proposed in the literature. However, most of them suffer from issues related to privacy, coverage or convenience. In the current paper, we propose an infrared array sensor-based approach to detect the presence/absence of a person in a room. We used a wide angle low resolution sensor (i.e., 32×24 pixels) to collect heat-related information from the area monitored, and used Deep Learning (DL) to identify the presence of up to 3 people with an accuracy reaching 97%. Our approach also detects of the presence or absence of a person with a 100% accuracy. Nevertheless, it allows identifying the location of the detected people within a room of dimensions 4×7.4 m with a margin of 0.3 m.

• Access to Document (DOI)

Joint Multiple Subspace-based BSS Method for Fetal Heart Rate Extraction from Non-invasive Recordings

Wang L., Ohtsuki T., Owada K., Honma N., Hayashi H.

Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS (Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS)  2020-July   616 - 620 2020.07

ISSN  9781728119908

　View Summary

© 2020 IEEE. Despite the enormous potential applications, non-invasive recordings have not yet made enough satisfaction for fetal disease detection. This is mainly due to the fetal ECG signal is contaminated by the maternal electrocardiograph (ECG) interference, muscle contractions, and motion artifacts. In this paper, we propose a joint multiple subspace-based blind source separation (BSS) approach to extract the fetal heart rate (HR), so that it could greatly reduce the effect of maternal ECG and motion artifacts. The approach relies on the estimation of the coefficient matrix formulated as the tensor decomposition in terms of multiple datasets. Since the objective function takes the coupling information from the stacking of the covariance matrix for multiple datasets into account, estimating the coefficient matrices is fulfilled not only on dependence across multiple datasets, but also can combine the extracted components across four different datasets. Numerical results demonstrate that the proposed method can achieve a high extracted HR accuracy for each dataset, when compared to some conventional methods.

• Access to Document (DOI)

Deep Reinforcement Learning for Economic Dispatch of Virtual Power Plant in Internet of Energy

Lin L., Guan X., Peng Y., Wang N., Maharjan S., Ohtsuki T.

IEEE Internet of Things Journal (IEEE Internet of Things Journal)  7 ( 7 ) 6288 - 6301 2020.07

　View Summary

© 2014 IEEE. With the high penetration of large-scale distributed renewable energy generation, the power system is facing enormous challenges in terms of the inherent uncertainty of power generation of renewable energy resources. In this regard, virtual power plants (VPPs) can play a crucial role in integrating a large number of distributed generation units (DGs) more effectively to improve the stability of the power systems. Due to the uncertainty and nonlinear characteristics of DGs, reliable economic dispatch in VPPs requires timely and reliable communication between DGs, and between the generation side and the load side. The online economic dispatch optimizes the cost of VPPs. In this article, we propose a deep reinforcement learning (DRL) algorithm for the optimal online economic dispatch strategy in VPPs. By utilizing DRL, our proposed algorithm reduced the computational complexity while also incorporating large and continuous state space due to the stochastic characteristics of distributed power generation. We further design an edge computing framework to handle the stochastic and large-state space characteristics of VPPs. The DRL-based real-time economic dispatch algorithm is executed online. We utilize real meteorological and load data to analyze and validate the performance of our proposed algorithm. The experimental results show that our proposed DRL-based algorithm can successfully learn the characteristics of DGs and industrial user demands. It can learn to choose actions to minimize the cost of VPPs. Compared with the deterministic policy gradient algorithm and DDPG, our proposed method has lower time complexity.

• Access to Document (DOI)

Online Recursive ICA Algorithm Used for Motor Imagery EEG Signal

Lin X., Wang L., Ohtsuki T.

Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS (Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS)  2020-July   502 - 505 2020.07

ISSN  9781728119908

　View Summary

© 2020 IEEE. Electroencephalogram (EEG) signals are important to study the activities of human brains. The independent component analysis (ICA) algorithm is a practical blind source separation (BSS) technique that can separate EEG sources from artifacts effectively. However, most traditional ICA algorithms assume that the mixing process is instantaneous and off-line. In this paper, a novel framework based on the extension of the online recursive ICA algorithm (ORICA) is proposed to apply for motor imagery (MI) EEG recording. The contributions are as follows. Firstly, we show ORICA's adaptability to accurate and effective source separation used for artifact-contaminated MI EEG recording. Secondly, to identify EOG signals on the output of source separation, the topographic map is presented to distinguish the target signals. The experimental results show that the proposed framework is able to be applied to process MI EEG recording in real-time situations.

• Access to Document (DOI)

Deep Clustering with LSTM for Vital Signs Separation in Contact-free Heart Rate Estimation

Ye C., Gui G., Ohtsuki T.

IEEE International Conference on Communications (IEEE International Conference on Communications)  2020-June 2020.06

ISSN  9781728150895

　View Summary

© 2020 IEEE. So far, most separation approaches of vital signs such as heartbeat and respiration, are implemented based on linear mixtures. However, some literatures have reported that non-linear mixtures actually occur in the associated applications, e.g., heart rate (HR) estimation with Doppler radar, where the simple linear demixing architecture may limit the effect of source separation. In addition, the human motions during HR measurement further complicate the mixing processes. The issue motivates us to exploit a more suitable separation approach to deal with contact-free HR estimation, considering non-linear mixtures including motions. A semi-supervised deep clustering (DC) is proposed to separate the three mixed sources of heartbeat, respiration, and motions, by segmenting the spectrogram of Doppler signal. First, through training a deep recurrent neural network (RNN) with long short-term memory (LSTM) via heartbeat/respiration-only data, the embeddings to each frame-sample from spectrogram can be acquired, which enables feature optimization in a lower dimensional space. Then, in the test phase, K-means clusters the embeddings associated with each source, to infer the masks used for spectrogram segmentation. The proposed deep clustering has three main strengths: It (i) gets rid of the restriction of mixture class, relying on data mining; (ii) can handle three-source mixtures by training two sorts of source-independent samples; (iii) only requires the mixtures from single-channel. The HR measurement experiments on subjects' sitting still and typing, validate the improvements of accuracy and robustness by our proposal, over some prevailing approaches in signal decomposition or separation.

• Access to Document (DOI)