Paper Info
Deep convolutional autoencoder applied for noise reduction in range-doppler maps of fmcw radars
M de Oliveira, M Bekooij
2020 IEEE International Radar Conference (RADAR), 630-635, 2020.
Reviews
From
yassendobrev, rated
4/5.
Oct. 29, 2021.
An interesting paper proposing to use a convolutional autoencoder (CAE) to replace the classical (OS-)CFAR / peak detection combination and demonstrating that it outperforms it. They use a combination of synthetic and augmented data to train their network. They labelled their dataset manually, but I wish they provided more information about the labeling process.
Deep convolutional autoencoder applied for noise reduction in range-doppler maps of fmcw radars
M de Oliveira, M Bekooij
2020 IEEE International Radar Conference (RADAR), 630-635, 2020.
Reviews
From
yassendobrev, rated
4/5.
Oct. 29, 2021.
An interesting paper proposing to use a convolutional autoencoder (CAE) to replace the classical (OS-)CFAR / peak detection combination and demonstrating that it outperforms it. They use a combination of synthetic and augmented data to train their network. They labelled their dataset manually, but I wish they provided more information about the labeling process.