Radar Image Reconstruction from Raw ADC Data using Parametric Variational Autoencoder with Domain Adaptation

Stephan, Michael; Stadelmayer, Thomas; Santra, Avik; Fischer, Georg; Weigel, Robert; Lurz, Fabian

doi:10.1109/icpr48806.2021.9412858

Cited by 14 publications

(4 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The bound in (11) can also be expressed in terms of empirical measures rather than true probability measures by the addition of Rademacher complexity terms [17,Chapter 3].…”

Section: A Margin Disparity Discrepancymentioning

confidence: 99%

“…Both supervised and unsupervised domain adaptation methods have already been investigated in the Radar-ML community to overcome several problems, including individual patient differences [8], aspect angle variations [9], syntheticto-real adaptation [10] or environmental differences [11]. In the case of cross-configuration adaptation, Khodabakhshandeh et al [12] use supervised techniques such as Few-shot Adversarial Domain Adaptation (FADA) [13] or domain adaptation using Stochastic Neighborhood Embedding (d-SNE) [14] to adapt their trained human activity classifier to new frequencymodulated continuous-wave (FMCW) radar setups using few data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Unsupervised Domain Adaptation across FMCW Radar Configurations Using Margin Disparity Discrepancy

Hernangómez¹,

Bjelaković²,

Servadei³

et al. 2022

Preprint

View full text Add to dashboard Cite

Commercial radar sensing is gaining relevance and machine learning algorithms constitute one of the key components that are enabling the spread of this radio technology into areas like surveillance or healthcare. However, radar datasets are still scarce and generalization cannot be yet achieved for all radar systems, environment conditions or design parameters. A certain degree of fine tuning is, therefore, usually required to deploy machine-learning-enabled radar applications. In this work, we consider the problem of unsupervised domain adaptation across radar configurations in the context of deep-learning human activity classification using frequency-modulated continuous-wave.For that, we focus on the theory-inspired technique of Margin Disparity Discrepancy, which has already been proved successful in the area of computer vision. Our experiments extend this technique to radar data, achieving a comparable accuracy to fewshot supervised approaches for the same classification problem.

show abstract

“…The bound in (11) can also be expressed in terms of empirical measures rather than true probability measures by the addition of Rademacher complexity terms [17,Chapter 3].…”

Section: A Margin Disparity Discrepancymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Unsupervised Domain Adaptation across FMCW Radar Configurations Using Margin Disparity Discrepancy

Hernangómez¹,

Bjelaković²,

Servadei³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The fast advances in Machine Learning (ML) research help to solve more and more important problems. Lately, sophisticated ML methods found their way into radar signal processing and analysis [1][2][3]. However, generating high-quality datasets that adequately represent reality to its full extent is still highly demanding.…”

Section: Introduction 1motivationmentioning

confidence: 99%

Improving Radar Human Activity Classification Using Synthetic Data with Image Transformation

Hernangómez

Visentin

Servadei

et al. 2022

Sensors

View full text Add to dashboard Cite

Machine Learning (ML) methods have become state of the art in radar signal processing, particularly for classification tasks (e.g., of different human activities). Radar classification can be tedious to implement, though, due to the limited size and diversity of the source dataset, i.e., the data measured once for initial training of the Machine Learning algorithms. In this work, we introduce the algorithm Radar Activity Classification with Perceptual Image Transformation (RACPIT), which increases the accuracy of human activity classification while lowering the dependency on limited source data. In doing so, we focus on the augmentation of the dataset by synthetic data. We use a human radar reflection model based on the captured motion of the test subjects performing activities in the source dataset, which we recorded with a video camera. As the synthetic data generated by this model still deviates too much from the original radar data, we implement an image transformation network to bring real data close to their synthetic counterpart. We leverage these artificially generated data to train a Convolutional Neural Network for activity classification. We found that by using our approach, the classification accuracy could be increased by up to 20%, without the need of collecting more real data.

show abstract

“…Data processing of radar data is a major topic in the research community with numerous publications for different purposes, reaching from gesture recognition [5] and air writing [1] to people counting [3] and human activity classification [9]. Even small scale movements like vital signs [2] can be detected by this technology.…”

Section: Introductionmentioning

confidence: 99%

Frequency Modulated Continuous Wave Radar-Based Navigation Algorithm using Artificial Neural Network for Autonomous Driving

Valtl

Issakov

2021

2021 IEEE International Intelligent Transportation Systems Conference (ITSC)

View full text Add to dashboard Cite

Autonomous driving is a highly complex task, which involves the use of numerous sensors and various algorithms. Testing of algorithms is difficult and therefore mostly done in simulations. Radar technology will play a key part due to various advantages. In this paper we present a solution to one aspect of autonomous driving, which is the development of a detection algorithm on a moving platform, which is capable of tracking and sending the commands to follow a preceding object, by means of sensor data from a low power 60 GHz Frequency Modulated Continuous Wave (FMCW) radar. The moving platform is based on a miniaturized autonomous vehicle that is used for data gathering as well as algorithm evaluation.To the best of the author's knowledge, this is the first time that processing of radar data via Deep Convolutional Neural Networks (DCNN) for navigation purposes is performed in real time on the edge device operating in a real world environment and not simulative.

show abstract

Radar Image Reconstruction from Raw ADC Data using Parametric Variational Autoencoder with Domain Adaptation

Cited by 14 publications

References 20 publications

Unsupervised Domain Adaptation across FMCW Radar Configurations Using Margin Disparity Discrepancy

Unsupervised Domain Adaptation across FMCW Radar Configurations Using Margin Disparity Discrepancy

Improving Radar Human Activity Classification Using Synthetic Data with Image Transformation

Frequency Modulated Continuous Wave Radar-Based Navigation Algorithm using Artificial Neural Network for Autonomous Driving

Contact Info

Product

Resources

About