Pith. sign in

REVIEW 2 major objections 5 minor 68 references

Deep Generative Models for Bayesian Inference on High-Rate Sensor Data: Applications in Automotive Radar and Medical Imaging

T0 review · 2 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A review of deep generative model techniques for Bayesian inverse problems in high-rate sensor data, covering structured noise, companding, model-based scores, and acceleration methods.

desk verdict A competent but heavily self-referential review of DGM-based Bayesian inference for radar and ultrasound; the synthesis is useful, but Eq (3.8) mixes domains and the core independence assumption needs a critical paragraph. read the letter →

arxiv 2504.12154 v1 pith:FZOEPIIY submitted 2025-04-16 eess.SP

Tristan S.W. Stevens , Jeroen Overdevest , Oisín Nolan , Wessel L. van Nierop , Ruud J.G. van Sloun , Yonina C. Eldar This is my paper · ORCID
classification eess.SP
keywords datagenerativebayesianmodelsapplicationsautomotivecontextdeep
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper is about using deep generative models, AI systems that learn what typical data looks like and can create new examples, to make sense of fast and messy sensor data from car radar and medical scanners. These sensors produce huge amounts of data very quickly, and the data often contains noise or interference that is hard to describe with simple math. The authors review ways to adapt generative models to this setting: learning the noise separately with its own model, compressing the signal range so models train better, using physics knowledge to guide reconstruction, and making the models faster by reusing past frames, unrolling algorithms, distilling smaller models, or actively choosing which measurements to take next.

The main examples are ultrasound imaging, where multipath reflections create a hazy look; automotive radar, where other cars' radar signals interfere; and MRI, where choosing which k-space lines to measure can speed up scans. The paper does not run new experiments. It describes methods from earlier papers, many by the same research group, and organizes them around two challenges: model mismatch, when the math model does not match reality, and real-time speed.

If the described methods work as claimed, they could eventually make medical imaging and self-driving car sensors faster and more reliable while using less data. But the evidence here is mostly taken from previous publications, and one key equation about companding appears to mix mathematical domains, so a careful reader should check the original sources before relying on the details.

Extended reading notes

Core claim

Central assertion: the direct application of natural-image generative models to raw sensor data is not straightforward, and the resulting challenges (high dynamic range, model mismatch, latency) can be addressed by the surveyed techniques: joint posterior sampling with learned structured-noise priors, companding of RF data, model-based score functions, temporal initialization, deep unfolding, and active compressed sensing. Stated in the abstract: 'we will discuss some of these challenges and offer approaches to address them, all in the context of high-rate real-time sensing applications in automotive radar and medical imaging.' If the paper is correct, these techniques extend DGM-based Bayesian posterior sampling to real-time automotive radar and medical ultrasound/MRI.

Load-bearing premise

The load-bearing premise is that structured noise n is statistically independent of the signal x. Introduced in Section 2(a) after Eq (2.1): 'Given the complexity of capturing potential dependencies between n and x, we assume independence, i.e., p(n|x) ≈ p(n) and instead learn the marginal distribution p(n) in a fully data-driven fashion.' This premise is required for the factorized joint posterior in Eq (3.1) and the coupled reverse diffusion in Eq (3.2). If multipath or interference depends on the scene x, the learned marginal prior p(n) is misspecified and the posterior samples are biased.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The manuscript is a review of deep generative model (DGM) based Bayesian inference for high-rate sensor data, with applications to automotive radar and medical imaging. It introduces a forward model y = Ax + n + ε, reviews score-based diffusion posterior sampling via Tweedie's formula, and then discusses three classes of methods for model mismatch: modeling structured noise with separate DGMs, companding to handle high dynamic range, and model-based score functions. It also reviews acceleration strategies (temporal initialization, deep unfolding, knowledge distillation) and active compressed sensing. The central claim is that these techniques make DGM-based posterior sampling applicable to real-time sensing, illustrated with ultrasound multipath dehazing, radar interference mitigation, and accelerated MRI.

Significance. The paper's background derivation of posterior sampling is clean and the survey of active compressed sensing covers several recent external methods, which gives it value as a structured introduction to the area. If the reviewed techniques are correct, they would indeed extend DGM-based Bayesian inference to high-rate sensing. However, the technical credibility of the main model-mismatch framework is undermined by a factorized joint posterior that relies on an independence assumption the paper itself contradicts for the ultrasound example, and by an algebraic error in the companding likelihood (Eq. (3.8)). These issues are fixable, but they are load-bearing for the paper's central claims.

major comments (2)
  1. [Section 2(a), Eq. (3.1), Box 3.1] The structured-noise framework assumes p(n|x) ≈ p(n) and learns only the marginal p(n), which is what justifies the factorized joint posterior in Eq. (3.1) and the coupled reverse-time SDEs in Eqs. (3.3)-(3.4). However, Section 2(a) itself states that in ultrasound multipath scattering, 'n is inherently a function of x', and Box 3.1 models clutter and multipath components as n generated by the same transmitted pulse. In this regime the independence assumption is misspecified: unpaired training of separate score models s_θ(x_τ) and s_ϕ(n_τ) captures only marginals, and the likelihood term must carry the full burden of enforcing consistency. The manuscript does not analyze the resulting bias in posterior samples, nor does it provide any calibration or validation study for exactly the multipath regime where the assumption is invoked. Since this framework is also used for radar interference mitigation, this concern is load-bearing for the claim that the surveyed techniques address model mismatch.
  2. [Section 3(b), Eq. (3.8)] The second equality in Eq. (3.8), C(A x_RF,0|τ + n_RF,0|τ) = C(C^{-1}(A x_{0|τ}) + C^{-1}(n_{0|τ})), is not correct. The companding functions C and C^{-1} act elementwise on RF-domain vectors, while A x_{0|τ} is a measurement-domain vector; applying C^{-1} to A x_{0|τ} is not a defined operation in this setup, and A and C^{-1} do not commute (A C^{-1}(x) ≠ C^{-1}(A x) in general). Moreover, if y in Eq. (3.7) is the raw RF measurement, the likelihood residual should be y - (A C^{-1}(x_{0|τ}) + C^{-1}(n_{0|τ})), not y - C(A C^{-1}(x_{0|τ}) + C^{-1}(n_{0|τ})). This invalidates the data consistency term used to produce the ultrasound results in Fig. 2 unless the variables are redefined and the derivation is corrected.
minor comments (5)
  1. [Section 3(b), Eq. (3.7)] The gradient on the right-hand side of Eq. (3.7) should be with respect to both x_τ and n_τ, not only x_τ; as written, the update for n_τ is undefined.
  2. [Section 3(b)] The notation x_RF,0|τ and x_{0|τ} is introduced without clearly stating whether these are Tweedie estimates in the RF domain or in the companded domain; please define both explicitly before Eq. (3.8).
  3. [Section 4(b), AdaSense paragraph] The text refers to 'principle components' of the posterior covariance; this should be 'principal components'.
  4. [Box 3.1] The phrase 'hundreds or thousands Gigabits per second' should be 'hundreds or thousands of gigabits per second'.
  5. [Section 4(a)(iii)] Knowledge distillation is described as a promising acceleration method, but no sensing-specific example or reference to high-rate sensing applications is given; the review should either provide one or explicitly state that this remains future work.
Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central review rests on the independence of structured noise from the signal, the Gaussian thermal noise model, and standard score-based diffusion results. No invented entities are introduced; the companding parameter and sparsity penalty are hand-chosen.

free parameters (2)
  • mu (companding parameter) = hand-chosen, e.g., 255 or 50
    Controls the degree of dynamic-range compression in Eq (3.6); adjusted per application so companded RF resembles image pixel intensities. Not fitted to data.
  • lambda_tau (sparsity penalty schedule) = time-step dependent, hand-tuned
    Used in the model-based score denoising step in Eq (3.10) for radar; selected per diffusion step, not inferred from data.
assumptions (5)
  • domain assumption Structured noise n is independent of the signal x, i.e., p(n|x) ≈ p(n)
    Stated in Section 2(a); required for factorized joint posterior p(x,n|y) = p(y|x,n)p(x)p(n) in Eq (3.1). Violated in multipath/diffraction scenarios where n depends on x.
  • domain assumption Thermal noise is additive white Gaussian with known variance σ²_ε
    Used in the forward model (2.1) and in likelihood terms throughout Sections 2 and 3; real sensor noise may be colored or heteroscedastic.
  • standard math Tweedie's formula gives the MMSE estimate x0|τ from the score
    Invoked in Eq (2.6) and used by DPS and ADS; standard result (Efron 2011), not proved in the paper.
  • standard math Reverse-time SDE with substituted score yields samples from the posterior
    Used in Eq (2.4) and (3.2); relies on score-based generative modeling theory (Song et al. 2020).
  • domain assumption Learned score functions sθ and sφ are accurate enough for posterior sampling
    All examples assume the trained diffusion networks approximate the true score well; no convergence or error bounds are given for finite data, HDR, or structured-noise settings.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Deep Generative Models for Bayesian Inference on High-Rate Sensor Data: Applications in Automotive Radar and Medical Imaging." pith.science (2026). https://pith.science/paper/FZOEPIIY

@misc{pith2026250412154,
  author       = {Pith},
  title        = {Pith review of: Deep Generative Models for Bayesian Inference on High-Rate Sensor Data: Applications in Automotive Radar and Medical Imaging},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FZOEPIIY}},
  note         = {Machine review of arXiv:2504.12154}
}
read the original abstract

Deep generative models have been studied and developed primarily in the context of natural images and computer vision. This has spurred the development of (Bayesian) methods that use these generative models for inverse problems in image restoration, such as denoising, inpainting, and super-resolution. In recent years, generative modeling for Bayesian inference on sensory data has also gained traction. Nevertheless, the direct application of generative modeling techniques initially designed for natural images on raw sensory data is not straightforward, requiring solutions that deal with high dynamic range signals acquired from multiple sensors or arrays of sensors that interfere with each other, and that typically acquire data at a very high rate. Moreover, the exact physical data-generating process is often complex or unknown. As a consequence, approximate models are used, resulting in discrepancies between model predictions and the observations that are non-Gaussian, in turn complicating the Bayesian inverse problem. Finally, sensor data is often used in real-time processing or decision-making systems, imposing stringent requirements on, e.g., latency and throughput. In this paper, we will discuss some of these challenges and offer approaches to address them, all in the context of high-rate real-time sensing applications in automotive radar and medical imaging.

Figures

Figures reproduced from arXiv: 2504.12154 by the authors.

Figure 1
Figure 1. Overview of the proposed joint posterior sampling method for removing structured [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Comparison between a structured noise DGM prior and a Gaussian prior for the task of dehazing in-vivo medical ultrasound data. Posterior estimates of the signal xˆ and noise (haze) nˆ are shown for each method, alongside corresponding gCNR [35] (↑) values, highlighting the improved performance of the structured noise prior. Figure adopted from [33]. (b) High dynamic range Unlike natural images, which typically have … view at source ↗
Figure 3
Figure 3. (Left) RF signals and their companded versions, where the [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: A semi-correlated interference scenario where a large portion of the raw data is [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Illustrative example of an active compressed sensing step for [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

68 extracted references · 58 canonical work pages

  1. [1]

    2018 fastMRI: An open dataset and benchmarks for accelerated MRI

    Zbontar J, Knoll F, Sriram A, Murrell T, Huang Z, Muckley MJ, Defazio A, Stern R, Johnson P, Bruno M et al.. 2018 fastMRI: An open dataset and benchmarks for accelerated MRI. arXiv preprint arXiv:1811.08839

    arXiv 2018

  2. [2]

    2019 Deep learning in ultrasound imaging.Proceedings of the IEEE 108, 11–29

    Van Sloun RJ, Cohen R, Eldar YC. 2019 Deep learning in ultrasound imaging.Proceedings of the IEEE 108, 11–29

    work page 2019

  3. [3]

    2022 Reframing Fast-Chirp FMCW Transceivers for Future Automotive Radar: The pathway to higher resolution.IEEE Solid-State Circuits Magazine 14, 44–55

    Doris K, Filippi A, Jansen F. 2022 Reframing Fast-Chirp FMCW Transceivers for Future Automotive Radar: The pathway to higher resolution.IEEE Solid-State Circuits Magazine 14, 44–55

    work page 2022

  4. [4]

    2022 Score-based diffusion models for accelerated MRI.Medical image analysis 80, 102479

    Chung H, Ye JC. 2022 Score-based diffusion models for accelerated MRI.Medical image analysis 80, 102479

    work page 2022

  5. [5]

    2022 High-resolution image synthesis with latent diffusion models

    Rombach R, Blattmann A, Lorenz D, Esser P, Ommer B. 2022 High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition pp. 10684–10695

    work page 2022

  6. [6]

    2024 The llama 3 herd of models.arXiv preprint arXiv:2407.21783

    Dubey A, Jauhri A, Pandey A, Kadian A, Al-Dahle A, Letman A, Mathur A, Schelten A, Yang A, Fan A et al.. 2024 The llama 3 herd of models.arXiv preprint arXiv:2407.21783

    arXiv 2024

  7. [7]

    2006 Near-optimal signal recovery from random projections: Universal encoding strategies?

    Candes EJ, Tao T. 2006 Near-optimal signal recovery from random projections: Universal encoding strategies?. IEEE transactions on information theory 52, 5406–5425

    work page 2006

  8. [8]

    2012 Compressed sensing: theory and applications

    Eldar YC, Kutyniok G. 2012 Compressed sensing: theory and applications . Cambridge university press

    work page 2012

Show all 68 references
  1. [9]

    2015Sampling theory: Beyond bandlimited systems

    Eldar YC. 2015Sampling theory: Beyond bandlimited systems . Cambridge University Press

  2. [10]

    2018 A systematic review of compressive sensing: Concepts, implementations and applications.IEEE access 6, 4875–4894

    Rani M, Dhok SB, Deshmukh RB. 2018 A systematic review of compressive sensing: Concepts, implementations and applications.IEEE access 6, 4875–4894

    2018

  3. [11]

    2014 Fourier-domain beamforming: the path to compressed ultrasound imaging

    Chernyakova T, Eldar YC. 2014 Fourier-domain beamforming: the path to compressed ultrasound imaging. IEEE transactions on ultrasonics, ferroelectrics, and frequency control 61, 1252–1267

    2014

  4. [12]

    2019 Strategic undersampling and recovery using compressed sensing for enhancing ultrasound image quality

    Ramkumar A, Thittai AK. 2019 Strategic undersampling and recovery using compressed sensing for enhancing ultrasound image quality. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 67, 547–556

    2019

  5. [13]

    2020 Learning sub-sampling and signal recovery with applications in ultrasound imaging.IEEE Transactions on Medical Imaging 39, 3955–3966

    Huijben IA, Veeling BS, Janse K, Mischi M, van Sloun RJ. 2020 Learning sub-sampling and signal recovery with applications in ultrasound imaging.IEEE Transactions on Medical Imaging 39, 3955–3966

    2020

  6. [14]

    2011 Robust recovery of synthetic aperture radar data from uniformly under-sampled measurements

    Nguyen LH, Tran TD. 2011 Robust recovery of synthetic aperture radar data from uniformly under-sampled measurements. In2011 IEEE International Geoscience and Remote Sensing Symposium pp. 3554–3557. IEEE. 16royalsocietypublishing.org/journal/rsta Phil. Trans. R. Soc. A 0000000

    2011

  7. [15]

    2018 Sub-Nyquist radar systems: Temporal, spectral, and spatial compression

    Cohen D, Eldar YC. 2018 Sub-Nyquist radar systems: Temporal, spectral, and spatial compression. IEEE Signal Processing Magazine 35, 35–58

    2018

  8. [16]

    2019 Compressed sensing MRI: a review from signal processing perspective.BMC Biomedical Engineering 1, 8

    Ye JC. 2019 Compressed sensing MRI: a review from signal processing perspective.BMC Biomedical Engineering 1, 8

    2019

  9. [17]

    2023 Model-Based Deep Learning

    Shlezinger N, Whang J, Eldar YC, Dimakis AG. 2023 Model-Based Deep Learning. Proceedings of the IEEE 111, 465–499

    2023

  10. [18]

    Gorban AN, Tyukin IY. 2018 Blessing of dimensionality: mathematical foundations of the statistical physics of data.Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 376, 20170237

    2018

  11. [19]

    2013 Auto-encoding variational bayes.arXiv preprint arXiv:1312.6114

    Kingma DP. 2013 Auto-encoding variational bayes.arXiv preprint arXiv:1312.6114

    2013 arXiv

  12. [20]

    2014 Generative adversarial nets

    Goodfellow I, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozair S, Courville A, Bengio Y. 2014 Generative adversarial nets. Advances in neural information processing systems 27

    2014

  13. [21]

    2018 Glow: Generative flow with invertible 1x1 convolutions

    Kingma DP, Dhariwal P. 2018 Glow: Generative flow with invertible 1x1 convolutions. Advances in neural information processing systems 31

    2018

  14. [22]

    2020 Score-Based Generative Modeling through Stochastic Differential Equations

    Song Y, Sohl-Dickstein J, Kingma DP, Kumar A, Ermon S, Poole B. 2020 Score-Based Generative Modeling through Stochastic Differential Equations. InInternational Conference on Learning Representations

    2020

  15. [23]

    Advances in neural information processing systems 32

    SongY,Ermon S.2019Generativemodeling byestimating gradientsofthe datadistribution. Advances in neural information processing systems 32

  16. [24]

    2024 A regularized conditional GAN for posterior sampling in image recovery problems.Advances in Neural Information Processing Systems 36

    Bendel M, Ahmad R, Schniter P. 2024 A regularized conditional GAN for posterior sampling in image recovery problems.Advances in Neural Information Processing Systems 36

    2024

  17. [25]

    2023 Pseudoinverse-guided diffusion models for inverse problems

    Song J, Vahdat A, Mardani M, Kautz J. 2023 Pseudoinverse-guided diffusion models for inverse problems. InInternational Conference on Learning Representations

    2023

  18. [26]

    2023 A variational perspective on solving inverse problems with diffusion models.arXiv preprint arXiv:2305.04391

    Mardani M, Song J, Kautz J, Vahdat A. 2023 A variational perspective on solving inverse problems with diffusion models.arXiv preprint arXiv:2305.04391

    2023 arXiv

  19. [27]

    2024 A survey on diffusion models for inverse problems.arXiv preprint arXiv:2410.00083

    Daras G, Chung H, Lai CH, Mitsufuji Y, Ye JC, Milanfar P, Dimakis AG, Delbracio M. 2024 A survey on diffusion models for inverse problems.arXiv preprint arXiv:2410.00083

    2024 arXiv

  20. [28]

    2022 Diffusion posterior sampling for general noisy inverse problems.arXiv preprint arXiv:2209.14687

    Chung H, Kim J, Mccann MT, Klasky ML, Ye JC. 2022 Diffusion posterior sampling for general noisy inverse problems.arXiv preprint arXiv:2209.14687

    2022 arXiv

  21. [29]

    2011 Tweedie’s formula and selection bias.Journal of the American Statistical Association 106, 1602–1614

    Efron B. 2011 Tweedie’s formula and selection bias.Journal of the American Statistical Association 106, 1602–1614

    2011

  22. [30]

    2024 Denoising: A powerful building-block for imaging, inverse problems, and machine learning.arXiv preprint arXiv:2409.06219

    Milanfar P, Delbracio M. 2024 Denoising: A powerful building-block for imaging, inverse problems, and machine learning.arXiv preprint arXiv:2409.06219

    2024 arXiv

  23. [31]

    2024 Model-Based Diffusion for Mitigating Automotive Radar Interference

    Overdevest J, Wei X, van Gorp H, van Sloun RJG. 2024 Model-Based Diffusion for Mitigating Automotive Radar Interference. In 2024 IEEE International Conference on Acoustics, Speech, and Signal Processing Workshops (ICASSPW) pp. 284–288

    2024

  24. [32]

    2024 Dehazing Ultrasound using Diffusion Models.IEEE Transactions on Medical Imaging pp

    Stevens TSW, Meral FC, Yu J, Apostolakis IZ, Robert JL, Van Sloun RJ. 2024 Dehazing Ultrasound using Diffusion Models.IEEE Transactions on Medical Imaging pp. 1–1

    2024

  25. [33]

    StevensTSW,vanGorpH,MeralFC,ShinJ,YuJ,RobertJL,vanSlounRJ.2025Removing Structured Noise using Diffusion Models.Transactions on Machine Learning Research

  26. [34]

    2021 Compressed ultrasound imaging: From sub-Nyquist rates to super resolution.IEEE BITS the information theory magazine 1, 27–44

    Drori O, Mamistvalov A, Solomon O, Eldar YC. 2021 Compressed ultrasound imaging: From sub-Nyquist rates to super resolution.IEEE BITS the information theory magazine 1, 27–44

    2021

  27. [35]

    2019 The generalized contrast-to-noise ratio: A formal definition for lesion detectability

    Rodriguez-Molares A, Rindal OMH, D’hooge J, Måsøy SE, Austeng A, Bell MAL, Torp H. 2019 The generalized contrast-to-noise ratio: A formal definition for lesion detectability. IEEE transactions on ultrasonics, ferroelectrics, and frequency control 67, 745–759

    2019

  28. [36]

    2025 Unlimited sampling beyond modulo

    Azar E, Mulleti S, Eldar YC. 2025 Unlimited sampling beyond modulo. Applied and Computational Harmonic Analysis 74, 101715

    2025

  29. [37]

    2021 Automotive Radar — From First Efforts to Future Systems

    Waldschmidt C, Hasch J, Menzel W. 2021 Automotive Radar — From First Efforts to Future Systems. IEEE Journal of Microwaves 1, 135–148

    2021

  30. [38]

    2016 Chirp diversity waveform design and detection by stretch processing

    Lulu A, Mobasseri BG. 2016 Chirp diversity waveform design and detection by stretch processing. In 2016 IEEE Radar Conference (RadarConf) pp. 1–6

    2016

  31. [39]

    1991 A stepped chirp technique for range resolution enhancement

    McGroary F, Lindell K. 1991 A stepped chirp technique for range resolution enhancement. In NTC ’91 - National Telesystems Conference Proceedings pp. 121–126

    1991

  32. [40]

    2017 Automotive radar interference mitigation using a sparse sampling approach

    Bechter J, Roos F, Rahman M, Waldschmidt C. 2017 Automotive radar interference mitigation using a sparse sampling approach. In2017 European Radar Conference (EURAD) pp. 90–93

    2017

  33. [41]

    2020 Automotive Radar Signal Interference Mitigation Using RNN with Self Attention

    Mun J, Ha S, Lee J. 2020 Automotive Radar Signal Interference Mitigation Using RNN with Self Attention. InICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) pp. 3802–3806. 17royalsocietypublishing.org/journal/rsta Phil. Trans. ...

    2020

  34. [42]

    2020 Automotive Radar Interference Mitigation using a Convolutional Autoencoder

    Fuchs J, Dubey A, Lübke M, Weigel R, Lurz F. 2020 Automotive Radar Interference Mitigation using a Convolutional Autoencoder. In 2020 IEEE International Radar Conference (RADAR) pp. 315–320

    2020

  35. [43]

    2021 Estimating the Magnitude and Phase of Automotive Radar Signals Under Multiple Interference Sources With Fully Convolutional Networks

    Ristea NC, Anghel A, Ionescu RT. 2021 Estimating the Magnitude and Phase of Automotive Radar Signals Under Multiple Interference Sources With Fully Convolutional Networks. IEEE Access 9, 153491–153507

    2021

  36. [44]

    2021 Complex-valued convolutional neural networks for enhanced radar signal denoising and interference mitigation

    Fuchs A, Rock J, Toth M, Meissner P, Pernkopf F. 2021 Complex-valued convolutional neural networks for enhanced radar signal denoising and interference mitigation. In2021 IEEE Radar Conference (RadarConf21) pp. 1–6. IEEE

    2021

  37. [45]

    2023 Signal Reconstruction for FMCW Radar Interference Mitigation Using Deep Unfolding

    Overdevest J, Koppelaar A, Bekooij M, Youn J, Sloun Rv. 2023 Signal Reconstruction for FMCW Radar Interference Mitigation Using Deep Unfolding. InICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) pp. 1–5

    2023

  38. [46]

    2024 Neurally Augmented Deep Unfolding for Automotive Radar Interference Mitigation.IEEE Transactions on Radar Systems 2, 712–724

    Overdevest J, Koppelaar AGC, Youn J, Wei X, Sloun RJGv. 2024 Neurally Augmented Deep Unfolding for Automotive Radar Interference Mitigation.IEEE Transactions on Radar Systems 2, 712–724

    2024

  39. [47]

    2025 Sequential Posterior Sampling with Diffusion Models

    Stevens TSW, Nolan O, Robert JL, Van Sloun RJ. 2025 Sequential Posterior Sampling with Diffusion Models. In 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Hyderabad, India

    2025

  40. [48]

    2010 Learning fast approximations of sparse coding

    Gregor K, LeCun Y. 2010 Learning fast approximations of sparse coding. InProceedings of the 27th International Conference on International Conference on Machine Learning ICML’10 p. 399–406 Madison, WI, USA. Omnipress

    2010

  41. [49]

    2023 Learning to sample: Data-driven sampling and reconstruction of FRI signals.IEEE Access 11, 71048–71062

    Mulleti S, Zhang H, Eldar YC. 2023 Learning to sample: Data-driven sampling and reconstruction of FRI signals.IEEE Access 11, 71048–71062

    2023

  42. [50]

    Li Y, Bar-Shira O, Monga V, Eldar YC. 2023 p. 53–86. InDeep Algorithm Unrolling for Biomedical Imaging, p. 53–86. Cambridge University Press

    2023

  43. [51]

    2022 Deep Unfolding With Normalizing Flow Priors for Inverse Problems.IEEE Transactions on Signal Processing 70, 2962–2971

    Wei X, van Gorp H, Gonzalez-Carabarin L, Freedman D, Eldar YC, van Sloun RJG. 2022 Deep Unfolding With Normalizing Flow Priors for Inverse Problems.IEEE Transactions on Signal Processing 70, 2962–2971

    2022

  44. [52]

    2017 Unrolled Generative Adversarial Networks

    Metz L, Poole B, Pfau D, Sohl-Dickstein J. 2017 Unrolled Generative Adversarial Networks. arXiv preprint arXiv:1611.02163

    2017 arXiv

  45. [53]

    2019 Compressing gans using knowledge distillation

    Aguinaldo A, Chiang PY, Gain A, Patil A, Pearson K, Feizi S. 2019 Compressing gans using knowledge distillation. arXiv preprint arXiv:1902.00159

    2019 arXiv

  46. [54]

    2022 Progressive distillation for fast sampling of diffusion models.arXiv preprint arXiv:2202.00512

    Salimans T, Ho J. 2022 Progressive distillation for fast sampling of diffusion models.arXiv preprint arXiv:2202.00512

    2022 arXiv

  47. [55]

    2021 Solving inverse problems in medical imaging with score-based generative models.arXiv preprint arXiv:2111.08005

    Song Y, Shen L, Xing L, Ermon S. 2021 Solving inverse problems in medical imaging with score-based generative models.arXiv preprint arXiv:2111.08005

    2021 arXiv

  48. [56]

    2019 Inverse GANs for accelerated MRI reconstruction

    Narnhofer D, Hammernik K, Knoll F, Pock T. 2019 Inverse GANs for accelerated MRI reconstruction. In Wavelets and Sparsity XVIII vol. 11138 pp. 381–392. SPIE

    2019

  49. [57]

    2015 Info-greedy sequential adaptive compressed sensing.IEEE Journal of selected topics in signal processing 9, 601–611

    Braun G, Pokutta S, Xie Y. 2015 Info-greedy sequential adaptive compressed sensing.IEEE Journal of selected topics in signal processing 9, 601–611

    2015

  50. [58]

    2021 Active deep probabilistic subsampling

    Van Gorp H, Huijben I, Veeling BS, Pezzotti N, Van Sloun RJ. 2021 Active deep probabilistic subsampling. In International Conference on Machine Learning pp. 10509–10518. PMLR

    2021

  51. [59]

    2020 Experimental design for MRI by greedy policy search

    Bakker T, van Hoof H, Welling M. 2020 Experimental design for MRI by greedy policy search. Advances in Neural Information Processing Systems 33, 18954–18966

    2020

  52. [60]

    2022 Accelerated intravascular ultrasound imaging using deep reinforcement learning

    Stevens TSW, Chennakeshava N, de Bruijn FJ, Pekař M, van Sloun RJ. 2022 Accelerated intravascular ultrasound imaging using deep reinforcement learning. InICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) pp. 1216–1220. IEEE

    2022

  53. [61]

    2020 Uncertainty-driven adaptive sampling via GANs

    Sanchez T, Krawczuk I, Sun Z, Cevher V. 2020 Uncertainty-driven adaptive sampling via GANs. In NeurIPS 2020 Workshop on Deep Learning and Inverse Problems

    2020

  54. [62]

    2023 Active Subsampling Using Deep Generative Models by Maximizing Expected Information Gain

    van de Camp KC, Joudeh H, Antunes DJ, van Sloun RJ. 2023 Active Subsampling Using Deep Generative Models by Maximizing Expected Information Gain. InICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) pp. 1–5. IEEE

    2023

  55. [63]

    2017 Improved training of wasserstein gans

    Gulrajani I, Ahmed F, Arjovsky M, Dumoulin V, Courville AC. 2017 Improved training of wasserstein gans. Advances in neural information processing systems 30

    2017

  56. [64]

    2007 Approximating the Kullback Leibler divergence between Gaussian mixture models

    Hershey JR, Olsen PA. 2007 Approximating the Kullback Leibler divergence between Gaussian mixture models. In 2007 IEEE International Conference on Acoustics, Speech and Signal Processing-ICASSP’07 vol. 4 pp. IV–317. IEEE. 18royalsocietypublishing.org/journal/rsta Phil. Trans. ...

    2007

  57. [65]

    1998 Gradient-based learning applied to document recognition

    LeCun Y, Bottou L, Bengio Y, Haffner P. 1998 Gradient-based learning applied to document recognition. Proceedings of the IEEE 86, 2278–2324

    1998

  58. [66]

    2024 Adaptive Compressed Sensing with Diffusion-Based Posterior Sampling

    Elata N, Michaeli T, Elad M. 2024 Adaptive Compressed Sensing with Diffusion-Based Posterior Sampling. arXiv preprint arXiv:2407.08256

    2024 arXiv

  59. [67]

    2022 Denoising diffusion restoration models.Advances in Neural Information Processing Systems 35, 23593–23606

    Kawar B, Elad M, Ermon S, Song J. 2022 Denoising diffusion restoration models.Advances in Neural Information Processing Systems 35, 23593–23606

    2022

  60. [68]

    2025 Active Diffusion Subsampling

    Nolan O, Stevens TSW, van Nierop WL, Sloun RV. 2025 Active Diffusion Subsampling. Transactions on Machine Learning Research

    2025

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.