Knowledge Exploitation for Human Micro-Doppler Classification

Karabacak, Cesur; Gürbüz, Sevgi Zübeyde; Gürbüz, Ali Cafer; Güldoǧan, Mehmet Burak; Hendeby, Gustaf; Gustafsson, Fredrik

doi:10.1109/lgrs.2015.2452311

Cited by 75 publications

(48 citation statements)

References 21 publications

(11 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1. (these physical interpretations of the envelopes have been validated by [7] using motion capture data). The following gait parameters were extracted using these envelopes.…”

Section: Doppler Radar Gait Measurementmentioning

confidence: 94%

Screening of mild cognitive impairment in elderly via Doppler radar gait measurement

2020

View full text Add to dashboard Cite

This letter presents a screening technique for elderly adults with mild cognitive impairment towards the early detection of dementia based on daily gait measurement using a Doppler radar. The gait parameters corresponding to walking speed, gait cycle, and leg velocities were remotely extracted using a simple Doppler radar system for elderly participants aged 65 years and above. The screening capabilities of the participants with mild cognitive impairment were investigated by using the extracted gait parameters. The results verified that our Doppler radar technique achieved mild cognitive impairment screening with approximately 94% sensitivity and 69% specificity.

show abstract

“…1. (these physical interpretations of the envelopes have been validated by [7] using motion capture data). The following gait parameters were extracted using these envelopes.…”

Section: Doppler Radar Gait Measurementmentioning

confidence: 94%

Screening of mild cognitive impairment in elderly via Doppler radar gait measurement

2020

View full text Add to dashboard Cite

show abstract

“…The features used as inputs of the proposed Bi-LSTM are metrics extracted from the original radar spectrograms and wearable data as a function of time. For the radar sensor, Doppler centroid and bandwidth are considered, together with upper and lower envelopes, mean, standard deviation, skewness, and kurtosis calculated for each time bin of the spectrogram [34], [35]. For the Wrist IMU, the data include 9 features with the X, Y, and Z axes data of accelerometer, gyroscope, and magnetometer.…”

Section: B Bi-lstm-based Deep Neural Networkmentioning

confidence: 99%

Bi-LSTM Network for Multimodal Continuous Human Activity Recognition and Fall Detection

et al. 2020

View full text Add to dashboard Cite

This paper presents a framework based on multi-layer bi-LSTM network (bidirectional Long Short-Term Memory) for multimodal sensor fusion to sense and classify daily activities' patterns and high-risk events such as falls. The data collected in this work are continuous activity streams from FMCW radar and three wearable inertial sensors on the wrist, waist, and ankle. Each activity has a variable duration in the data stream so that the transitions between activities can happen at random times within the stream, without resorting to conventional fixed-duration snapshots. The proposed bi-LSTM implements soft feature fusion between wearable sensors and radar data, as well as two robust hard-fusion methods using the confusion matrices of both sensors. A novel hybrid fusion scheme is then proposed to combine soft and hard fusion to push the classification performances to approximately 96% accuracy in identifying continuous activities and fall events. These fusion schemes implemented with the proposed bi-LSTM network are compared with conventional sliding window approach, and all are validated with realistic "leaving one participant out" (L1PO) method (i.e. testing subjects unknown to the classifier). The developed hybrid-fusion approach is capable of stabilizing the classification performance among different participants in terms of reducing accuracy variance of up to 18.1% and increasing minimum, worst-case accuracy up to 16.2%.

show abstract

“…Radar is a kind of microwave sensor which can exploit signal processing techniques for target detecting, imaging and classification [1][2][3][4][5][6][7] . When the microwave pulse transmit by the radar bounced from the moving targets, the carrier frequency of the signal will be shifted, which is known as the Doppler effect.…”

Section: Introductionmentioning

confidence: 99%

“…Different kinds of targets may have different kinds of micro-motions, which makes micro-Doppler signature classification a promising means for identifying radar targets. The early studies are mainly focused on template-based classification of many kinds of targets, such as aircraft 1,[16][17] , human aquatic activity [3][4][5][6][7]18 , missile defence [19][20][21] and ground vehicles 10,[22][23][24] . Chen analysed the micro-Doppler modulation with a mathematical model in detail and introduced the micro-Doppler effect in radar [8][9] .…”

Section: Introductionmentioning

confidence: 99%

Template free Micro Doppler Signature Classification for Wheeled and Tracked Vehicles

et al. 2019

View full text Add to dashboard Cite

The micro-Doppler signature is a time-varying frequency modulation imparted on radar echo caused by target's micro-motion. To save the trouble of constructing template in the target classification, this paper investigates the micro-Doppler signature of wheeled and tracked vehicles and proposes a template-free classification method. Firstly, the echo signature is established and the micro-Doppler difference of these two kinds of targets is analysed. Secondly, some new micro-Doppler features are defined according to their difference. The new defined features are micro-Doppler bandwidth, micro-Doppler expansion rate and micro-Doppler peak number. According to the characteristic of the micro-Doppler in the time-frequency domain, we proposed to realise the feature extraction by Hough transformation. Lastly, template-free subjection functions are proposed to define the relationship between the features and the vehicles. By fuzzy comprehensive evaluation, the final classification result is obtained by combining the subjection probabilities together. Experimental results based on the simulated data and measured data are presented, which prove that the algorithm has good performance.

show abstract

Knowledge Exploitation for Human Micro-Doppler Classification

Cited by 75 publications

References 21 publications

Screening of mild cognitive impairment in elderly via Doppler radar gait measurement

Screening of mild cognitive impairment in elderly via Doppler radar gait measurement

Bi-LSTM Network for Multimodal Continuous Human Activity Recognition and Fall Detection

Template free Micro Doppler Signature Classification for Wheeled and Tracked Vehicles

Contact Info

Product

Resources

About