8 results
Now showing1 - 8 of 8
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Light-weight and person-independent radar-based hand gesture recognition for classification and regression of continuous gestures(IEEE, 2024-05-01); ; ; ; This paper proposes a novel pre-processing technique for radar-based short-range gesture sensing using a frequency modulated continuous wave (FMCW) radar. The pre-processing is light-weight and works without Fourier transformation. The signal after pre-processing represents the backscattering central dynamics of the hand as a complex-valued time signal of a point target. It is shown that the proposed processing provides competitive classification results compared to conventional frequency domain-based solutions, while being less computationally intensive and having better generalization performance. The pre-processed time domain signal preserves a high temporal resolution of the hand movement. Due to this fact, it is possible to integrate a periodic control gesture into the system. In doing so, the system not only detects that a gesture is performed continuously and periodically, but also estimates its speed. This is an essential property for controlling scalable parameters such as brightness or volume at different speeds. The real-time capability was proven on a Raspberry Pi 3B with an ARM Cortex-A53 CPU. The proposed processing causes a CPU utilization of only 6%. The nn inference is done within 75 ms with a classification accuracy of 96.7%.Publicationtype: Journal Article38 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Improved indoor semi-static human target detection and localization using FMCW radarThis paper presents a novel processing approach for semi-static human indoor perception and localization with Frequency modulated continuous wave (FMCW) radar. When analyzing semi-static targets with FMCW radar, minimal phase variations of standing or sitting human targets in slow-time can result in missed detections under traditional processing schemes. The common approach of studying micro-Doppler signatures over an extended period of time can, however, be costly in terms of both memory and compute resources. In this work, we introduce an approach with low memory and computational footprint, which relies on signatures beyond micro-Doppler for a novel presence score. We show on real-world recordings how this maximal variance based presence score can bring improvement in detection of static humans. To further increase detections in the range angle space, we introduce a novel angle of arrival estimation by creating a synthetic spatial signal from all phase differences between antennas over time. We show a 22.9% gain in human detection performance on a large dataset of humans in semi-static applications and highlight the reduction in computational complexity of both of these concepts.Publicationtype: Conference Paper22 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Physically-Interpretable Data Augmentation for Multi-Range Hand Gesture Recognition Using FMCW Radar Time SeriesIn this paper, a robust ac hgr system using a ac fmcw radar and InceptionTime networks on data augmented time series is implemented. The paper proposes multiple data augmentation techniques for radar-based ac hgr. Since a realistic manipulation of raw radar data frames or even range-Doppler maps is a very complex challenge, we instead propose data manipulation on physically interpretable time series of range, azimuth and elevation angles extracted from the data. Due to working on physically interpretable time series data, we can on the one hand make use of well explored existing augmentation techniques for time series and on the other hand do use-case specific interpretable augmentations, such as simulating a different aspect angle or range. To investigate the system, a data recording process covering multiple ranges, angles and types of gestures is carried out. The gain in accuracy from the proposed data augmentation scheme amounts to more than 4 percentage points to reach a global prediction accuracy higher than 95% for a very diverse dataset.Publicationtype: Journal Article33 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Out-of-distribution detection for radar-based gesture recognition using metric-learningThe paper addresses the question how and to what extent metric learning can be beneficial for reducing the false alarm rate in radar-based hand gesture recognition systems. To this end, we evaluate different metric learning approaches for out-of-distribution or unknown motion detection. We found that metric learning can help to significantly increase the out-of-distribution capabilities of the network. We further investigated what conditions must be met for metric learning to work well, and found that the composition of the data set for known gestures has a large influence on the out-of-distribution detection rate.Publicationtype: Conference Paper52 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Radar-Based Gesture Recognition Using a Variational Autoencoder With Deep Statistical Metric Learning(2022-11-01); ; ; Radar-based gesture recognition systems are a promising concept for new human–machine interfaces. However, the systems are sensitive to user-dependent gesture patterns, sensor noise characteristics, background environments, and unknown gestures or background motions. To overcome such challenges, we propose a variational autoencoder architecture-based deep metric learning, which is optimized using a novel loss function combining a statistical distance triplet loss and the center loss. By learning the statistical distance between distributions, the proposed solution, compared with triplet loss, is able to better learn the nonlinear characteristics in the data, is less sensitivity to training strategy, and is capable of creating close knit embedding clusters. Moreover, it is experimentally demonstrated that the proposed gesture recognition system using radar spectrograms has improved classification accuracy and random gesture rejection accuracy compared with the state-of-the-art deep metric learning approaches. Furthermore, we proof the real-world capability of the system by implementing a real-time demonstrator, which allows playing “Tetris” with hand gestures.Publicationtype: Journal Article74 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Data-driven Radar Processing Using a Parametric Convolutional Neural Network for Human Activity Classification(2021-09-01); ; ; The paper proposes a data-driven pre-processing optimization for radar data using a parametric convolutional neural network. The proposed method is applied on human activity classification as a use case. Present radar-based activity recognition system exploit micro-Doppler signature by generating Doppler spectrograms or a temporal series of range-Doppler maps, followed by deep neural networks or machine learning approaches for classification. Those radar data representations are typically generated on the basis of short-time Fourier transformations. A Fourier transformation equally resolves the frequency space, which may be sub-optimal in some applications. Although deep convolutional neural networks (DCNN) have been shown to implicitly learn features from raw sensor data in other fields, such as speech recognition, yet, for the case of radar-based DCNNs, pre-processing is required to develop a scalable and robust classification or regression application. In this paper, we propose a parametric convolutional neural network that mimics the radar pre-processing across fast-time and slow-time radar data through 2D sinc filter or 2D wavelet filter kernels to extract features for classification of various human activities. During training only the filter parameters of the 2D sinc filters or 2D wavelets are learned, leading to optimized feature representation for the classification task. It is demonstrated that our proposed solution shows improved results compared to equivalent DCNN architectures that rely on Doppler spectrograms or radar data cubes as input data.Publicationtype: Journal Article84 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Improved Target Detection and Feature Extraction using a Complex-Valued Adaptive Sine Filter on Radar Time Domain DataIn state-of-art radar classification tasks the raw time domain ADC data is transformed to frequency domain wherein the target is detected. Feature data, such as the micro-Doppler signature, is extracted for classification of e.g. a performed gesture or a person based on its gait. However, the transformation into frequency domain using short-time Fourier transform resulting in erroneous target detection due to superposition of target components thus leading to sub-optimal feature for subsequent classification task. Thus, in this paper we propose a target feature extraction approach that operates directly on 2D time domain radar data by using a complex-valued adaptive 2D sine filter. The proposed approach tracks the target's slow-time and fast-time frequencies by a regulation loop, which progressively adjusts the filter's center frequency to the target location. We demonstrate that the proposed approach extracts better features in both single- and multi-target scenarios, leading to improved classification accuracy in gait classification problems using radars. Furthermore, the proposed time domain feature extraction approach facilitates the use of a parametric neural network that works directly on time domain data.Publicationtype: Conference Paper54 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Radar image reconstruction from Raw ADC data using parametric variational autoencoder with domain adaptationThis paper presents a parametric variational autoencoder-based human target detection and localization framework working directly with the raw analog-to-digital converter data from the frequency modulated continuous wave radar. We propose a parametrically constrained variational autoencoder, with residual and skip connections, capable of generating the clustered and localized target detections on the range-angle image. Furthermore, to circumvent the problem of training the proposed neural network on all possible scenarios using real radar data, we propose domain adaptation strategies whereby we first train the neural network using ray tracing based model data and then adapt the network to work on real sensor data. This strategy ensures better generalization and scalability of the proposed neural network even though it is trained with limited radar data. We demonstrate the superior detection and localization performance of our proposed solution compared to the conventional signal processing pipeline and earlier state-of-art deep U-Net architecture with range-doppler images as inputs.Publicationtype: Conference Paper55