Appel Choose France
L’appel Choose France est ouvert jusqu’au 31 mars. C’est une belle opportunité pour attirer en France des talents...
23 Janvier 2024
Catégorie : Stagiaire
Location: Ecole d’ingénieurs de numérique (Isep), Paris, France
Starting date: March 2024 (negotiable if necessary)
Supervisors: Idowu AJAYI and Wafa NJIMA
General Presentation / Technological Context
The internship is centered on millimeter-wave(mmWave) Frequency-Modulated Continuous Wave (FMCW) radars. The mmWave radar can accurately measure the distance, azimuth, and velocity of a target thanks to its small size, low cost, and ability to be effective in most weather conditions. This is why it is a significant technology within the autonomous driving community [1]. However, with the rapid growth in the number of automotive radar sensors and bandwidth limitations, the challenge of interference has become a major issue in radar systems [2].
Interference suppression is the process of eliminating unwanted signals from desired information signal. Legacy approaches used signal processing techniques but with the recent advancements in deep learning, it has become a method of choice for suppressing interference [3]. Similarly in radar systems, it has been proven that deep learning-based methods have the tendency to outperform classical methods.
Some of the many deep learning techniques that have recently been explored are CNN-based [4], [5], RNN-based methods [6], Autoencoder-based methods [7], etc. Despite the promises of these AI- based methods, a major challenge in this area remains the scarcity and quality of dataset. Due to the stochastic nature of wireless communication, it is indeed difficult to get a dataset that will be optimal for generalization in different realizations. Moreso, when we consider interference suppression, it becomes more challenging to have labelled data, where we capture real interference and the ground truth.
A significant contribution in this domain is the publicly accessible dataset provided by [1]. Different indoor and outdoor scenarios were considered, consisting of both RDG-D (color and depth) data along with their initial measurement unit (IMU) details. The internship will start with this data set and preprocess with various interference scenarios. This will be the basis on which the deep learning interference suppression models will be proposed.
Work Organization / Expected Deliverables
This internship has the following objectives:
Qualifications and Required Skills
This internship has a minimum 5 months duration beginning in March 2024. Internships will be awarded on a rolling basis and candidates are encouraged to apply early.
Good level in oral and writing English (French optional).
How to Apply
Interested candidates should send a detailed CV, a one-page motivation letter, two academic references and M1 transcript to idowu.ajayi@isep.fr and wafa.njima@isep.fr. For applications, the subject of your email should be " Internship Application - DL for Interference Suppression in Radar".
Closing date: 11:59 pm 31st January 2024 (GMT+1 Time Zone)
Interviews will be conducted by videoconference between 2-3 February 2024.
References
[1] T.-Y.Lim,S.A.Markowitz,andM.N.Do,“RaDICaL:ASynchronizedFMCWRadar,Depth,IMUand RGB Camera Data Dataset With Low-Level FMCW Radar Signals,” IEEE J. Sel. Top. Signal Process., vol. 15, no. 4, pp. 941–953, Jun. 2021, doi: 10.1109/JSTSP.2021.3061270.
[2] L.Liu,R.Guan,F.Ma,J.Smith,andY.Yue,“Radar-STDA:AHigh-PerformanceSpatial-Temporal Denoising Autoencoder for Interference Mitigation of FMCW Radars.” arXiv, Jul. 18, 2023. doi: 10.48550/arXiv.2307.09063.
[3] T.Oyedare,V.K.Shah,D.J.Jakubisin,andJ.H.Reed,“InterferenceSuppressionUsingDeep Learning: Current Approaches and Open Challenges,” IEEE Access, vol. 10, pp. 66238–66266, 2022, doi: 10.1109/ACCESS.2022.3185124.
[4] N.-C.Ristea,A.Anghel,andR.T.Ionescu,“FullyConvolutionalNeuralNetworksforAutomotive Radar Interference Mitigation,” in 2020 IEEE 92nd Vehicular Technology Conference (VTC2020- Fall), Nov. 2020, pp. 1–5. doi: 10.1109/VTC2020-Fall49728.2020.9348690.
[5] J.Rock,M.Toth,P.Meissner,andF.Pernkopf,“CNNsforInterferenceMitigationandDenoisingin Automotive Radar Using Real-World Data”.
[6] J.Mun,H.Kim,andJ.Lee,“ADeepLearningApproachforAutomotiveRadarInterference Mitigation,” in 2018 IEEE 88th Vehicular Technology Conference (VTC-Fall), Aug. 2018, pp. 1–5. doi: 10.1109/VTCFall.2018.8690848.
[7] M.L.L.deOliveiraandM.J.G.Bekooij,“DeepConvolutionalAutoencoderAppliedforNoise Reduction in Range-Doppler Maps of FMCW Radars,” in 2020 IEEE International Radar Conference (RADAR), Apr. 2020, pp. 630–635. doi: 10.1109/RADAR42522.2020.9114719.