Radar-Based Human Activity Recognition Under the Limited Measurement Data Support Using Domain Translation

Yang, Yang; Zhang, Yutong; Ji, Haoran; Li, Beichen; Song, Chunying

doi:10.1109/lsp.2022.3207948

Cited by 5 publications

(3 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, the authors of [33] proposed a rotational shift method to augment radar point cloud data. Recently, a two-stage domain adaptation scheme was presented in [34] to address the lack of training data for radar-based HAR systems. For data augmentation, they used a GAN-based domain-translation network to translate the simulated spectrograms into measurement-like spectrograms with the help of small measurement datasets.…”

Section: Related Workmentioning

confidence: 99%

A Simulation-Based Framework for the Design of Human Activity Recognition Systems Using Radar Sensors

Waqar,

Pätzold

2024

IEEE Internet Things J.

View full text Add to dashboard Cite

Modern human activity recognition (HAR) systems are designed using large amounts of experimental data. So far, real-data-driven or experimental-based HAR systems using Wi-Fi or radar systems have shown considerable results. However, the acquisition of large, clean, and labeled training datasets remains a crucial impediment to the progress of experimentalbased HAR systems. Therefore, in this paper, a paradigm shift from the experimental to a fully simulation-based design of HAR systems is proposed in the context of radar sensing. An endto-end simulation framework is proposed as a proof-of-concept that can simulate realistic millimeter-wave radar signatures for synthesized human motion. We designed a human motion synthesis tool that emulates different types of human activities and generates the spatial trajectories accordingly. These trajectories are processed by a geometric model with respect to user-defined antenna configurations. Considering the long-and short-time stationarity of wireless channels, we synthesize the raw in-phase and quadrature data and process the data to simulate the radar signatures for emulated human activities. Finally, a simulated and a real HAR dataset were used to train and test a simulation-based HAR system, respectively, which gave an average (maximum) classification accuracy of 94% (98.4%). The main advantage of the proposed simulation framework is that the training effort for radar-based classifiers, e.g., gesture recognition systems, can be minimized drastically.

show abstract

Section: Related Workmentioning

confidence: 99%

A Simulation-Based Framework for the Design of Human Activity Recognition Systems Using Radar Sensors

Waqar,

Pätzold

2024

IEEE Internet Things J.

View full text Add to dashboard Cite

show abstract

“…Recently, the issue of limited radar data is also discussed in [101] where a two-stage domain adaptation approach is provided for a generalized solution. Few samples are collected with OTS radar whereas few samples are generated using MCOAP dataset.…”

Section: H Miscellaneous Classification Approachesmentioning

confidence: 99%

Machine Learning for Healthcare Radars: Recent Progresses in Human Vital Sign Measurement and Activity Recognition

Ahmed,

Cho

2024

IEEE Commun. Surv. Tutorials

View full text Add to dashboard Cite

The unprecedented non-contact, non-invasive, and privacy-preserving nature of radar sensors has enabled various healthcare applications, including vital sign monitoring, fall detection, gait analysis, activity recognition, fitness evaluation, and sleep monitoring. Machine learning (ML) is revolutionizing every domain, with radar-based healthcare being no exception. Progress in the field of healthcare radars and ML is complementing the existing radar-based healthcare industry. This article provides an overview of ML usage for two major healthcare applications: vital sign monitoring and activity recognition. Vital sign monitoring is the most promising healthcare application of radar, as it can predict several chronic cardiac and respiratory diseases. Activity recognition is also a prominent application since the inability to perform activities may result in critical suffering. The article presents an overview of commercial radars, radar hardware, and historical progress of healthcare radars, followed by the usage of ML for healthcare radars. Subsequently, the paper discusses how ML can overcome the limitations of conventional radar data processing chains for healthcare radars. The article also touches upon recent generative ML concepts used in healthcare radars. Among several interesting findings, it was discovered that ML does not completely replace existing vital sign monitoring algorithms; rather, ML is deployed to overcome the limitations of traditional algorithms. On the other hand, activity recognition always relies on ML approaches. The most widely used algorithms for both applications are Convolutional Neural Network (CNN) followed by Support Vector Machine (SVM). Generative AI has the capability to augment data and is expected to have a significant impact soon. Recent trends, lessons learned from these trends, and future directions for both healthcare applications are presented in detail. Finally, the future work section discusses a wide range of healthcare topics for humans, ranging from neonates to elderly individuals.

show abstract

“…The use of the fewshot learning method was suggested by the authors of [27], which offers a unique way of augmenting the capabilities of pre-trained and pre-existing HAR systems. According to a recent study [28], a two-stage domain adaptation approach can be used to alleviate the data scarcity issue as well. With this approach, the simulated micro-Doppler signatures can be translated into measurement-like micro-Doppler signatures by using small real datasets.…”

Section: B Related Workmentioning

confidence: 99%

A Paradigm Shift From an Experimental-Based to a Simulation-Based Framework Using Motion-Capture Driven MIMO Radar Data Synthesis

Waqar,

Muaaz,

Sigg

et al. 2024

IEEE Sensors J.

View full text Add to dashboard Cite

The development of radar-based classifiers driven by empirical data can be highly demanding and expensive due to the unavailability of radar data. In this paper, we introduce an innovative simulation-based approach that addresses the data scarcity problem, particularly for our multiple-input multipleoutput (MIMO) radar-based direction-independent human activity recognition (HAR) system. To simulate realistic MIMO radar signatures, we first synthesize human motion and generate corresponding spatial trajectories. From these trajectories, a received radio frequency (RF) signal is synthesized using our MIMO channel model, which considers the non-stationary behavior of human motion and the multipath components originating from the scatterers on human body segments. Subsequently, the synthesized RF signals are processed to simulate MIMO radar signatures for various human activities. The proposed simulation-based direction-independent HAR system achieves a classification accuracy of 97.83% when tested with real MIMO radar data. A significant advantage of our simulation-based framework lies in its ability to facilitate multi-stage data augmentation techniques at the motion-layer, physical-layer, and signal-layer syntheses. This capability significantly reduces the training workload for radar-based classifiers. Importantly, our simulation-based proof-of-concept is applicable to single-input single-output and MIMO radars in monostatic, bistatic, and multistatic configurations, making it a versatile solution for realizing other radar-based classifiers, such as gesture classifiers.

show abstract

Radar-Based Human Activity Recognition Under the Limited Measurement Data Support Using Domain Translation

Cited by 5 publications

References 21 publications

A Simulation-Based Framework for the Design of Human Activity Recognition Systems Using Radar Sensors

A Simulation-Based Framework for the Design of Human Activity Recognition Systems Using Radar Sensors

Machine Learning for Healthcare Radars: Recent Progresses in Human Vital Sign Measurement and Activity Recognition

A Paradigm Shift From an Experimental-Based to a Simulation-Based Framework Using Motion-Capture Driven MIMO Radar Data Synthesis

Contact Info

Product

Resources

About