arXiv subjects

Kimitaka Sumi

Publications and source records attributed to Kimitaka Sumi.

5 recordsLinked to original sources

arXiv2026

Physiology-Informed Multivariate Variational Mode Decomposition for Orientation-Robust Sensing Using Distributed Millimeter-Wave Radar Systems

This study proposes a robust non-contact physiological sensing method using distributed radar systems to address the challenges arising from varying body orientations. Conventional physiological sensing methods using a single radar system suffer from performance degradation caused by variations in the relative geometry between the radar and the human body, particularly those caused by changes in body orientation. To help radar systems overcome this limitation, we propose a multivariate physiological variational mode decomposition method that extracts common respiratory and heart rates across multiple displacements of various body parts acquired from distributed radar systems. The proposed method incorporates harmonic constraints and a gap component tailored to the intrinsic nature of physiological signals into the model, thereby improving both estimation accuracy and robustness. Simultaneous measurements using four radar systems were performed on six participants under different participant-position and body-orientation conditions. Using the proposed method, we achieved success rates exceeding 90% for respiration and heartbeat estimation, which represent improvements of 19.8 and 25.3 percentage points, respectively, over rates using a conventional method based on a single radar system. Furthermore, we conducted simultaneous measurements of 16 participants using two radar systems in a multi-person scenario and successfully estimated both respiratory and heart rates with over 85% success rates. Our work contributes to the realization of practical radar-based non-contact physiological sensing by establishing a data fusion framework for distributed radar systems.

eess.SP↗

arXiv2026

Deformation-Aware Observation Modeling for Radar-Based Human Sensing via 3D Scan-Depth Sequence Fusion

Non-contact radar-based human sensing is often interpreted using simplified motion assumptions. However, respiration induces non-rigid surface deformation of the human body that impacts electromagnetic wave scattering and can degrade the robustness of measurements. To address this, we propose a surface-deformation-aware observation model for radar-based human sensing that fuses static high-resolution three-dimensional scanner measurements with temporal depth camera data to represent time-varying human surface geometry. Non-rigid registration using the coherent point drift algorithm is employed to align a static template with dynamic depth frames. Frame-wise electromagnetic scattering is subsequently computed using the physical optics approximation, allowing the reconstruction of intermediate-frequency radar signals that emulate radar observations. Validation against experimental radar data demonstrated that the proposed model exhibited greater robustness than a depth-sequence-only model under low-signal-quality conditions involving complex surface dynamics and multiple reflective sites. For two participants, the proposed model achieved higher Pearson correlation coefficients of 0.943 and 0.887 between model-derived and experimentally measured displacement waveforms, compared with 0.868 and 0.796 for the depth-sequence-only model. Furthermore, in a favorable case characterized by a single relatively-stationary reflective site, the proposed method achieved a correlation coefficient of 0.789 between model-derived and experimentally measured in-phase-quadrature magnitude variations. These results suggest that our sensor-fusion-based deformation-aware observation modeling can realistically reproduce radar observations and provide physically grounded insights into the interpretation of radar measurement variations.

eess.SP↗

arXiv2025

Simulation for Noncontact Radar-Based Physiological Sensing Using Depth-Camera-Derived Human 3D Model with Electromagnetic Scattering Analysis

This study proposes a method for simulating signals received by frequency-modulated continuous-wave radar during respiratory monitoring, using human body geometry and displacement data acquired via a depth camera. Unlike previous studies that rely on simplified models of body geometry or displacement, the proposed approach models high-frequency scattering centers based on realistic depth-camera-measured body shapes and motions. Experiments were conducted with six participants under varying conditions, including varying target distances, seating orientations, and radar types, with simultaneous acquisition from the radar and depth camera. Relative to conventional model-based methods, the proposed technique achieved improvements of 7.5%, 58.2%, and 3.2% in the correlation coefficients of radar images, displacements, and spectrograms, respectively. This work contributes to the generation of radar-based physiological datasets through simulation and enhances our understanding of factors affecting the accuracy of non-contact sensing.

eess.SP↗

arXiv2024

Accurate Radar-Based Heartbeat Measurement Using Higher Harmonic Components

This study proposes a radar-based heartbeat measurement method that uses the absolute value of the second derivative of the complex radar signal, rather than its phase, and the variational mode extraction method, which is a type of mode decomposition algorithm. We show that the proposed second-derivative-based approach can amplify the heartbeat component in radar signals effectively and also confirm that use of the variational mode extraction method represents an efficient way to emphasize the heartbeat component amplified via the second-derivative-based approach. We demonstrate estimation of the heart interbeat intervals using the proposed approach in combination with the topology method, which is an accurate interbeat interval estimation method. The performance of the proposed method is evaluated quantitatively using data obtained from eleven participants that were measured using a millimeter-wave radar system. When compared with conventional methods based on the phase of the complex radar signal, our proposed method can achieve higher accuracy when estimating the heart interbeat intervals; the correlation coefficient for the proposed method was increased by 0.20 and the root-mean-square error decreased by 23%.

eess.SP↗

arXiv2024

Complex Number Assignment in the Topology Method for Heartbeat Interval Estimation Using Millimeter-Wave Radar

The topology method is an algorithm for accurate estimation of instantaneous heartbeat intervals using millimeter-wave radar signals. In this model, feature points are extracted from the skin displacement waveforms generated by heartbeats and a complex number is assigned to each feature point. However, these numbers have been assigned empirically and without solid justification. This study used a simplified model of displacement waveforms to predict the optimal choice of the complex number assignments to feature points corresponding to inflection points, and the validity of these numbers was confirmed using analysis of a publicly available dataset.

eess.SP↗