FMCW Multi-Person Action Recognition System Based on Point Cloud Nearest Neighbor Sampling Algorithm

Sun, Mingxiu; Xu, Zhimeng; Sun, Beichen; Zhang, Shanshan

doi:10.1109/prai53619.2021.9551097

Cited by 6 publications

(4 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traditional target detection methods exploit spatial domain clustering algorithms to identify TPs and NPs in a radar point cloud based on the spatial Euclidean distance or its variants [19]. To classify fall, sit, and walk motions of two persons, the GMM is employed in [20] for target detection with mmWave radar point cloud as a preprocessing step. It assumes that the point cloud of each target follows a Gaussian distribution with specific mean and variance.…”

Section: Related Workmentioning

confidence: 99%

“…To reduce the negative impact of sparse TP distribution, some of the existing works introduce the temporal dimension for target detection or segmentation [20,32]. Specifically, in [32], multiple frames containing point clouds are formed into a greater point cloud while the newly formed point cloud is considered as a whole for target segmentation.…”

Section: Related Workmentioning

confidence: 99%

“…The performance of the multi-frame fusion mechanism is validated through experimental measured radar point cloud compared to the single-frame processing. Similarly, in [20], to address the issue of sparse TP clouds, a fixed number of frames are accumulated and process at the same time to improve the motion recognition of multiple targets. All these works are essentially accumulating the point cloud to a degree where the TP distribution becomes relatively dense enough until it can be correctly identified.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Recursive spatial-temporal clustering-based target detection with millimeter-wave radar point cloud

Gao

Wang³

et al. 2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Millimeter-wave (mmWave) radar plays a vital role in a wide range of applications such as security surveillance and environmental monitoring. This work investigates target detection with radar point cloud measurements in the slow-motion scenario. In contrast to the existing spatial domain clustering-based target detection methods, we adopt a recursive spatial-temporal clustering (STC)-based method to detect targets in the spatial and temporal domains jointly. Specifically, the points belonging to targets are obtained by clustering with a distance metric defined in the spatial-temporal domain. In addition, to ensure the feasibility of the proposed method for practical real-time implementation, a speed-up scheme that intends to reduce the computational complexity induced by clustering in both spatial and temporal dimensions is developed. We demonstrate the efficacy of the proposed recursive STC-based method through experimental mmWave radar point cloud data where multiple people walk simultaneously in an open space. The proposed method achieves decent target detection performance improvement compared to a widely-used clustering method for target detection while its computation time is negligible compared to radar data reception time.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Recursive spatial-temporal clustering-based target detection with millimeter-wave radar point cloud

Gao

Wang³

et al. 2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…Typically, FMCW radar sensors can extract human point cloud data [15] or 2-D domain spectra [16] to determine the target's activities. However, using the multi-frame point cloud accumulation method has limitations because some activities may have very similar point cloud results, such as sitting down and standing up, which are opposite activities.…”

Section: Introductionmentioning

confidence: 99%

Human Activity Recognition Method Based on FMCW Radar Sensor with Multi-Domain Feature Attention Fusion Network

Lin

Song

Zhao

et al. 2023

Sensors

View full text Add to dashboard Cite

This paper proposes a human activity recognition (HAR) method for frequency-modulated continuous wave (FMCW) radar sensors. The method utilizes a multi-domain feature attention fusion network (MFAFN) model that addresses the limitation of relying on a single range or velocity feature to describe human activity. Specifically, the network fuses time-Doppler (TD) and time-range (TR) maps of human activities, resulting in a more comprehensive representation of the activities being performed. In the feature fusion phase, the multi-feature attention fusion module (MAFM) combines features of different depth levels by introducing a channel attention mechanism. Additionally, a multi-classification focus loss (MFL) function is applied to classify confusable samples. The experimental results demonstrate that the proposed method achieves 97.58% recognition accuracy on the dataset provided by the University of Glasgow, UK. Compared to existing HAR methods for the same dataset, the proposed method showed an improvement of about 0.9–5.5%, especially in the classification of confusable activities, showing an improvement of up to 18.33%.

show abstract

Human Behavior Analysis Using Radar Data: A Survey

Miazek,

Żmudzińska,

karczmarek

et al. 2024

IEEE Access

View full text Add to dashboard Cite

FMCW Multi-Person Action Recognition System Based on Point Cloud Nearest Neighbor Sampling Algorithm

Cited by 6 publications

References 10 publications

Recursive spatial-temporal clustering-based target detection with millimeter-wave radar point cloud

Recursive spatial-temporal clustering-based target detection with millimeter-wave radar point cloud

Human Activity Recognition Method Based on FMCW Radar Sensor with Multi-Domain Feature Attention Fusion Network

Human Behavior Analysis Using Radar Data: A Survey

Contact Info

Product

Resources

About