Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

RadarSFD: Single-Frame Diffusion with Pretrained Priors for Radar Point Clouds

T0 review · 4 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read A single radar frame can be turned into a LiDAR-like point cloud by transferring pretrained depth priors into a latent diffusion model.

desk verdict Plausible and well-engineered single-frame radar-to-LiDAR diffusion, but the SOTA claim rests on quoted baselines and confounded ablations. read the letter →

arxiv 2509.18068 v2 pith:CEQO6YTZ submitted 2025-09-22 cs.RO eess.SP

Bin Zhao , Nakul Garg This is my paper · ORCID
classification cs.ROeess.SP
keywords single-frameradarlatentdiffusionradar-to-LiDARtranslationpretrainedpriorsBEVconditioningpointcloudreconstructionmmWaveChamferdistance
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

This paper tries to establish that a single millimeter-wave radar capture, without synthetic aperture or temporal stacking, can be converted into a dense LiDAR-like point cloud accurate enough for robotic perception. The authors argue that instead of learning scene geometry from scratch, a latent diffusion model can borrow structural priors from a pretrained monocular depth estimator and anchor them to the radar input by concatenating radar and LiDAR latents. On the RadarHD benchmark, the resulting system reduces Chamfer Distance from 56 cm (single-frame baseline) to 35 cm and Modified Hausdorff Distance from 45 cm to 28 cm, staying competitive with methods that use 5 to 41 frames. If true, this would make compact, low-power platforms like inspection drones and wearables viable for dense spatial sensing in fog, smoke, or dust.

What carries the argument

The central machinery is the conditional latent diffusion pipeline: a frozen Tiny AutoEncoder (TAESD) compresses radar and LiDAR bird's-eye-view images into 4-channel latent vectors; a denoising U-Net initialized from Marigold (a pretrained monocular depth estimator) performs iterative denoising with the radar latent concatenated channel-wise to the noisy LiDAR latent; and a dual-space objective combines the standard latent noise-prediction loss with pixel-space L1, SSIM, and LPIPS losses. The concatenation provides explicit spatial alignment, the Marigold init injects geometric priors about scene structure, and the pixel-space losses tether the output to the true LiDAR geometry, mitigating

What would settle it

Run all compared methods on the same test split at the same BEV resolution (256x512) with the same evaluation code and measure CD/MHD; if RadarSFD's CD is not below RAL'24's and the single-frame RadarHD baseline's, the state-of-the-art claim is refuted. Alternatively, a new-environment test with a different radar sensor would show whether the pretrained priors truly generalize or just memorize indoor corridors.

Watch

Extended reading notes

Core claim

The central claim is that a conditional latent diffusion model, initialized from a pretrained monocular depth estimator and conditioned by channel-wise concatenation of radar BEV latents with noisy LiDAR latents, can reconstruct dense, sharp LiDAR-like point clouds from a single radar frame without any motion or SAR. The authors report state-of-the-art results on RadarHD: 35 cm Chamfer Distance and 28 cm Modified Hausdorff Distance, an 8% and 3% improvement over the single-frame RAL'24 diffusion baseline, and competitive with multi-frame approaches. The key to this is that the pretrained depth priors supply a "world model" of walls, corners, and boundaries, while the radar input tethers the

Load-bearing premise

The central claim that single-frame radar can match multi-frame methods rests on comparing numbers reported by prior work under potentially different preprocessing (128x128 vs 256x512 BEVs), so if those numbers are not directly comparable, the measured improvement may not hold.

Editorial extensions

If this is right

  • Single-frame, no-SAR radar becomes a practical perception mode for size-, weight-, and power-constrained robots.
  • The pretrained-prior transfer recipe can be applied to other cross-modal sensor translations where paired data is scarce.
  • The dual-space objective reduces hallucination, making diffusion outputs more faithful to the conditioning input.
  • At 1.3 s per frame, the latent-space approach is faster than pixel-space diffusion (2.4 s per frame) while maintaining full 4 cm radar resolution.
  • Qualitative results show recovery of fine walls and narrow gaps that prior single-frame methods miss.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: the success of Marigold priors suggests that other pretrained geometry models (e.g., normal-estimation or surface-reconstruction diffusion models) could be swapped in to improve performance, even if this paper does not test them.
  • Editorial extension: if L1-only training matches the full dual-space loss on the reported metrics, a simplified objective may be sufficient for deployment, saving training cost without sacrificing measured accuracy.
  • Editorial extension: the single-frame capability opens an empirical test of whether the same method works outdoors with a different radar sensor, or whether the pretrained indoor priors fail in open-sky scenes.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes RadarSFD, a conditional latent diffusion model that reconstructs dense LiDAR-like BEV point clouds from a single millimeter-wave radar frame, without synthetic aperture or multi-frame stacking. The method encodes radar and LiDAR BEVs with a frozen TAESD VAE, concatenates the radar latent with the noisy LiDAR latent, initializes the denoising U-Net from Marigold, and trains with a combined latent-space noise-prediction loss plus pixel-space L1/SSIM/LPIPS losses. Evaluation on the RadarHD test split reports 0.35 m Chamfer Distance and 0.28 m Modified Hausdorff Distance, claiming state-of-the-art single-frame performance over the quoted RAL'24 baseline (0.38/0.29) and substantial gains over single-frame RadarHD (0.56/0.45). Ablations examine input representation, pretrained initialization, conditioning strategy, and loss composition.

Significance. If the quantitative claims are verified, the paper makes a useful contribution: it shows that pretrained monocular-depth diffusion priors can be transferred to radar-to-LiDAR translation, enabling a practical single-frame, no-SAR pipeline for SWaP-constrained robots. The use of a frozen VAE and latent diffusion for efficiency is sensible, and the qualitative results in Figures 1, 2, and 6 suggest a genuine improvement in structural fidelity over the single-frame baseline. The paper is empirical rather than theoretical; its strengths are the clear pipeline and the demonstration of cross-modal prior transfer. However, the headline 'state-of-the-art' claim depends on baseline numbers quoted from RAL'24 under a different BEV resolution and evaluation protocol, and the ablations that support the design attributions are partially confounded. These issues need to be resolved before the central claims can be accepted.

major comments (4)
  1. [§IV-D, Table II] The state-of-the-art claim rests on a 0.03 m CD and 0.01 m MHD margin over RAL'24, but the RAL'24 numbers are quoted from the original paper rather than produced under RadarSFD's evaluation protocol. Section IV-D itself states that RAL'24 operates on 128×128 BEVs while RadarSFD reconstructs at 256×512. Since CD and MHD are computed on extracted 2D point clouds, the grid resolution, BEV thresholding, and point-extraction procedure can shift the metrics independently of reconstruction quality. No error bars, seeds, or significance tests are reported for these margins. To support the claimed 8% CD improvement and the SOTA headline, the authors should rerun RAL'24 (and ideally RadarHD and ICRA'24 where possible) under a common preprocessing and evaluation pipeline, and report variance with paired statistical tests.
  2. [§IV-E, Table III] The ablation table changes two variables simultaneously in key rows. The 'Raw I/Q input' row uses cross-attention while the default uses concatenation, so the 2.3× CD degradation cannot be attributed solely to the input representation. Similarly, the 'Alt. pretraining (SDv2)' row uses cross-attention, conflating the choice of pretrained backbone with the conditioning strategy. The text in §IV-E2 claims that concatenation outperforms cross-attention across both I/Q inputs and SDv2 priors, but the table does not include the single-variable comparisons needed to support that claim. Please add rows that isolate each factor: e.g., raw I/Q with concatenation, SDv2 with concatenation, and optionally BEV light-threshold with cross-attention.
  3. [§IV-E3, Table III] The claim that the complete dual-space objective offers the best performance is not supported by the reported numbers. The 'L1 only' row matches the full system exactly on both metrics (0.35/0.28), while 'L1 + SSIM' is worse (0.42/0.34) and 'L1 + LPIPS' matches the full system. Thus the table provides no evidence that adding SSIM and LPIPS to L1 improves CD or MHD; in fact SSIM appears harmful on these metrics. If the authors wish to claim that the full loss is best, they need additional evidence, such as perceptual/structure metrics on the generated images, different weightings, or a statistical comparison across multiple runs. Otherwise the claim should be tempered to say that L1 is the primary driver and the additional losses were not observed to improve these particular metrics.
  4. [Abstract and §IV-D] The abstract states that 'experiments across new environments confirm strong generalization,' but the only supporting evidence is the qualitative Figure 2 for a single unseen building. No quantitative metrics are reported for the unseen-environment trajectories in the RadarHD test split, even though the dataset section describes three generalization conditions. Please provide quantitative results (CD/MHD, or per-condition breakdown) for unseen environments, or soften the generalization claim to qualitative observation.
minor comments (4)
  1. [Throughout] Typographical and formatting issues: 'generatitve' in Section I, 'halucination' in Section I, 'V AE' spacing throughout, 'we also prior work’s insight' missing a verb, and 'generalizablity' in Section II-A. These should be cleaned up.
  2. [§III, §IV] Reproducibility details are missing: no code repository, no training hyperparameters (loss weights λ_p, λ_L1, λ_SSIM, λ_LPIPS, learning rate, batch size, number of timesteps, inference steps, sampler), and no description of the point-cloud extraction procedure used to compute CD/MHD. These are needed for the 'practical reproducible recipe' claimed in the contributions.
  3. [Figure 7] The box plots for ablated runs would be more informative with sample sizes and statistical tests, especially since Table III reports only means and the text notes high variance for random initialization.
  4. [Table III] The row labeled 'Zero-threshold BEV' has different names in the first two columns ('Zero-threshold BEV' vs. 'BEV (zero-thresh)'). Please standardize for clarity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: RadarSFD is an empirical supervised learning pipeline; no claim reduces by construction to its inputs.

full rationale

RadarSFD is a conditional latent diffusion model trained on supervised radar–LiDAR pairs from RadarHD. The central claim is an empirical performance result, not a derivation from fitted parameters. The pretrained Marigold U-Net and frozen TAESD VAE are external priors transferred into the model; their use is not circular because they are not fit to the RadarHD test set and their geometric priors are learned from monocular depth estimation. The training objective (Eq. 5–7) combines a standard noise-prediction loss with pixel-space L1/SSIM/LPIPS losses, all of which are standard supervised losses; evaluation with Chamfer Distance and Modified Hausdorff Distance on held-out LiDAR ground truth is a conventional protocol, not a prediction forced by construction. The only concerns are comparability of quoted RAL'24 baselines (different BEV resolution, no re-run under identical protocol) and a mildly confounded ablation (input representation and conditioning strategy change together in two rows), but these are experimental-validity issues, not circularity. No self-citations are load-bearing, and no uniqueness theorem or ansatz is smuggled in via citation. Therefore, no circular step can be exhibited.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The method reuses off-the-shelf components (TAESD VAE, Marigold U-Net, diffusion losses) and does not introduce new physical entities. Free parameters (threshold, loss weights, timesteps) are not reported, making reproduction harder. The main postulated assumption is the transferability of monocular depth priors to radar geometry.

free parameters (3)
  • Radar BEV threshold (5% of magnitude) = 5%
    Chosen by hand to preserve sidelobes; affects input representation and is not justified by a principled optimization.
  • Loss weights (lambda_p, lambda_L1, lambda_SSIM, lambda_LPIPS) = not reported
    Weights in the dual-space objective, balanced empirically but values are not stated, hindering reproduction.
  • Diffusion timesteps and inference steps = not reported
    Training and sampling schedules are not specified, which are essential for a diffusion model.
assumptions (5)
  • standard math Diffusion models learn to reverse a fixed Gaussian corruption process (Eq. 1-2)
    Background from Ho et al. 2020, used as the framework for the method.
  • standard math A pretrained VAE encoder-decoder maps images to a lower-dimensional latent space and back (Eq. 3)
    Background from Rombach et al. 2022, used to compress BEVs.
  • domain assumption Radar BEV images generated from range/azimuth FFT with light thresholding retain sufficient scene geometry for dense reconstruction
    Motivated by RadarHD [3]; the paper relies on this without deriving an information-theoretic bound; the 5% threshold is an ad hoc choice.
  • domain assumption Geometric priors learned from monocular RGB depth estimation (Marigold) transfer to radar-to-LiDAR translation
    Central hypothesis; only tested empirically via ablation, not justified by theory.
  • domain assumption The RadarHD test split is representative and LiDAR ground truth is accurate
    Evaluation relies on this dataset; no cross-dataset quantitative evaluation beyond a qualitative figure (Fig. 2).

how reviews work

0 comments
Cite this review

Pith. "Pith review of RadarSFD: Single-Frame Diffusion with Pretrained Priors for Radar Point Clouds." pith.science (2026). https://pith.science/paper/CEQO6YTZ

@misc{pith2026250918068,
  author       = {Pith},
  title        = {Pith review of: RadarSFD: Single-Frame Diffusion with Pretrained Priors for Radar Point Clouds},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CEQO6YTZ}},
  note         = {Machine review of arXiv:2509.18068}
}
read the original abstract

Millimeter-wave radar provides robust perception in fog, smoke, dust, and low light, making it attractive for size-, weight-, and power-constrained robotic platforms. Existing radar imaging methods typically rely on synthetic aperture or multi-frame aggregation to improve resolution, which is impractical for small aerial, inspection, or wearable systems. We present RadarSFD, a conditional latent diffusion framework that reconstructs dense LiDAR-like point clouds from a single radar frame without motion or SAR. Our approach transfers geometric priors from a pretrained monocular depth estimator into the diffusion backbone, anchors them to radar inputs via channel-wise latent concatenation, and regularizes outputs with a dual-space objective combining latent and pixel-space losses. On the RadarHD benchmark, RadarSFD achieves state-of-the-art performance against baseline models. Qualitative results show recovery of fine walls and narrow gaps, and experiments across new environments confirm strong generalization. Ablation studies highlight the importance of pretrained initialization, radar BEV conditioning, and the dual-space loss. Together, these results establish a practical single-frame, no-SAR mmWave radar pipeline for dense point cloud perception in compact robotic systems.

Figures

Figures reproduced from arXiv: 2509.18068 by the authors.

Figure 1
Figure 1. Comparing radar-to-LiDAR point cloud reconstruc [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Real world test for generalization with completely unseen data in our campus building. All models are trained on the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Overview of the RadarSFD’s diffusion architecture. The radar and LiDAR images are first encoded into latent space. The radar latent c is concatenated with the noisy Li￾DAR latent zt and fed into a pretrained U-Net denoiser. The U-Net iteratively removes noise while preserving structure guided by the radar condition. After denoising, the decoder reconstructs a LiDAR-like point cloud with sharp geometry from a single … view at source ↗
Figures from the paper (3 more)
Figure 5
Figure 5. Figure 5: CDFs of reconstruction error (CD, MHD). RadarSFD achieves the steepest curves and lowest errors, outperform￾ing RAL’24, RadarHD single-frame, and CFAR, with most samples under 0.5 m CD and 0.4 m MHD. Both diffusion-based methods improve in the single-frame setting: RAL…
Figure 6
Figure 6. Figure 6: Qualitative comparison of point cloud reconstructions on four representative scenes with varying complexity. (a) [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Ablation box plots (Chamfer Distance). Left to right: input representation, pretrained priors, training losses. [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Dense Soft Weighting for Radar Ego-Velocity Estimation

    cs.RO 2026-07 conditional novelty 6.0 of 10

    Dense soft weighting of all range-Doppler cells yields training-free radar ego-velocity and covariance that cuts fused pose error 31–45% versus CFAR point-cloud baselines under a shared ESKF.

Reference graph

Works this paper leans on

36 extracted references · 2 linked inside Pith · cited by 1 Pith paper

  1. [1]

    Through fog high-resolution imaging using millimeter wave radar,

    J. Guan, S. Madani, S. Jog, S. Gupta, and H. Hassanieh, “Through fog high-resolution imaging using millimeter wave radar,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition, 2020, pp. 11 464–11 473

    2020

  2. [2]

    M. A. Richardset al.,Fundamentals of radar signal processing. Mcgraw-hill New York, 2005, vol. 1. Fig. 7: Ablation box plots (Chamfer Distance). Left to right: input representation, pretrained priors, training losses. Thresholded BEV inputs outperform raw I/Q; depth-pretrained priors (Marigold, SDv2) beat random initialization; adding pixel-space L1 drive...

    2005

  3. [3]

    High resolution point clouds from mmwave radar,

    A. Prabhakara, T. Jin, A. Das, G. Bhatt, L. Kumari, E. Soltanaghaei, J. Bilmes, S. Kumar, and A. Rowe, “High resolution point clouds from mmwave radar,”arXiv preprint arXiv:2206.09273, 2022

    arXiv 2022

  4. [4]

    Diffusion- based point cloud super-resolution for mmwave radar data,

    K. Luan, C. Shi, N. Wang, Y . Cheng, H. Lu, and X. Chen, “Diffusion- based point cloud super-resolution for mmwave radar data,” in2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024, pp. 11 171–11 177

    2024

  5. [5]

    Towards dense and accurate radar perception via efficient cross-modal diffusion model,

    R. Zhang, D. Xue, Y . Wang, R. Geng, and F. Gao, “Towards dense and accurate radar perception via efficient cross-modal diffusion model,” IEEE Robotics and Automation Letters, 2024

    2024

  6. [6]

    Repurposing diffusion-based image generators for monocular depth estimation,

    B. Ke, A. Obukhov, S. Huang, N. Metzger, R. C. Daudt, and K. Schindler, “Repurposing diffusion-based image generators for monocular depth estimation,” inProceedings of the IEEE/CVF con- ference on computer vision and pattern recognition, 2024, pp. 9492– 9502

    2024

  7. [7]

    Single-step latent diffusion for underwater image restoration,

    J. Wu, T. Wang, M. A. B. Siddique, M. J. Islam, C. Fermuller, Y . Aloimonos, and C. A. Metzler, “Single-step latent diffusion for underwater image restoration,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    2025

  8. [8]

    Image quality assessment: from error visibility to structural similarity,

    Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P. Simoncelli, “Image quality assessment: from error visibility to structural similarity,”IEEE transactions on image processing, vol. 13, no. 4, pp. 600–612, 2004

    2004

Show all 36 references
  1. [9]

    The unreasonable effectiveness of deep features as a perceptual metric,

    R. Zhang, P. Isola, A. A. Efros, E. Shechtman, and O. Wang, “The unreasonable effectiveness of deep features as a perceptual metric,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 586–595

    2018

  2. [10]

    Para- metric correspondence and chamfer matching: Two new techniques for image matching,

    H. G. Barrow, J. M. Tenenbaum, R. C. Bolles, and H. C. Wolf, “Para- metric correspondence and chamfer matching: Two new techniques for image matching,” Tech. Rep., 1977

    1977

  3. [11]

    A modified hausdorff distance for object matching,

    M.-P. Dubuisson and A. K. Jain, “A modified hausdorff distance for object matching,” inProceedings of 12th international conference on pattern recognition, vol. 1. IEEE, 1994, pp. 566–568

    1994

  4. [12]

    Fundamental limitations in antennas,

    R. C. Hansen, “Fundamental limitations in antennas,”Proceedings of the IEEE, vol. 69, no. 2, pp. 170–182, 2005

    2005

  5. [13]

    High-resolution frequency-wavenumber spectrum analysis,

    J. Capon, “High-resolution frequency-wavenumber spectrum analysis,” Proceedings of the IEEE, vol. 57, no. 8, pp. 1408–1418, 2005

    2005

  6. [14]

    Esprit-estimation of signal parameters via rotational invariance techniques,

    R. Roy and T. Kailath, “Esprit-estimation of signal parameters via rotational invariance techniques,”IEEE Transactions on acoustics, speech, and signal processing, vol. 37, no. 7, pp. 984–995, 2002

    2002

  7. [15]

    Multiple emitter location and signal parameter estima- tion,

    R. Schmidt, “Multiple emitter location and signal parameter estima- tion,”IEEE transactions on antennas and propagation, vol. 34, no. 3, pp. 276–280, 1986

    1986

  8. [16]

    Cfar detection for multiple target situations,

    M. Barkat, S. Himonas, and P. Varshney, “Cfar detection for multiple target situations,” inIEE Proceedings F (Radar and Signal Process- ing), vol. 136, no. 5. IET, 1989, pp. 193–209

    1989

  9. [17]

    Cfar: the principles of automatic radar detection in clutter,

    G. Minkler and J. Minkler, “Cfar: the principles of automatic radar detection in clutter,”Nasa sti/recon technical report a, vol. 90, p. 23371, 1990

    1990

  10. [18]

    Analysis of cfar processors in nonhomogeneous background,

    P. P. Gandhi and S. A. Kassam, “Analysis of cfar processors in nonhomogeneous background,”IEEE Transactions on Aerospace and Electronic systems, vol. 24, no. 4, pp. 427–445, 2002

    2002

  11. [19]

    Radar cfar thresholding in clutter and multiple target situations,

    H. Rohling, “Radar cfar thresholding in clutter and multiple target situations,”IEEE transactions on aerospace and electronic systems, no. 4, pp. 608–621, 2007

    2007

  12. [20]

    Near-field mimo-sar millimeter-wave imaging with sparsely sampled aperture data,

    M. E. Yanik and M. Torlak, “Near-field mimo-sar millimeter-wave imaging with sparsely sampled aperture data,”Ieee Access, vol. 7, pp. 31 801–31 819, 2019

    2019

  13. [21]

    Fast- fourier time-domain sar reconstruction for millimeter-wave fmcw 3-d imaging,

    A. V . Muppala, A. Y . Nashashibi, E. Afshari, and K. Sarabandi, “Fast- fourier time-domain sar reconstruction for millimeter-wave fmcw 3-d imaging,”IEEE Transactions on Microwave Theory and Techniques, vol. 72, no. 12, pp. 7028–7038, 2024

    2024

  14. [22]

    3d point cloud generation with millimeter-wave radar,

    K. Qian, Z. He, and X. Zhang, “3d point cloud generation with millimeter-wave radar,”Proceedings of the ACM on Interactive, Mo- bile, Wearable and Ubiquitous Technologies, vol. 4, no. 4, pp. 1–23, 2020

    2020

  15. [23]

    Radcloud: Real-time high-resolution point cloud generation using low-cost radars for aerial and ground vehicles,

    D. Hunt, S. Luo, A. Khazraei, X. Zhang, S. Hallyburton, T. Chen, and M. Pajic, “Radcloud: Real-time high-resolution point cloud generation using low-cost radars for aerial and ground vehicles,” in2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024,...

    2024

  16. [24]

    Mishape: Accurate human silhouettes and body joints from commodity millimeter-wave devices,

    A. Adhikari, H. Regmi, S. Sur, and S. Nelakuditi, “Mishape: Accurate human silhouettes and body joints from commodity millimeter-wave devices,”Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 6, no. 3, pp. 1–31, 2022

    2022

  17. [25]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,”Advances in neural information processing systems, vol. 33, pp. 6840–6851, 2020

    2020

  18. [26]

    High-resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High-resolution image synthesis with latent diffusion models,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 10 684–10 695

    2022

  19. [27]

    Adding conditional control to text-to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding conditional control to text-to-image diffusion models,” inProceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 3836–3847

    2023

  20. [28]

    Elucidating the design space of diffusion-based generative models,

    T. Karras, M. Aittala, T. Aila, and S. Laine, “Elucidating the design space of diffusion-based generative models,”Advances in neural information processing systems, vol. 35, pp. 26 565–26 577, 2022

    2022

  21. [29]

    Consistency models,

    Y . Song, P. Dhariwal, M. Chen, and I. Sutskever, “Consistency models,” 2023

    2023

  22. [30]

    The oxford radar robotcar dataset: A radar extension to the oxford robotcar dataset,

    D. Barnes, M. Gadd, P. Murcutt, P. Newman, and I. Posner, “The oxford radar robotcar dataset: A radar extension to the oxford robotcar dataset,” in2020 IEEE international conference on robotics and automation (ICRA). IEEE, 2020, pp. 6433–6438

    2020

  23. [31]

    nuscenes: A multimodal dataset for autonomous driving,

    H. Caesar, V . Bankiti, A. H. Lang, S. V ora, V . E. Liong, Q. Xu, A. Krishnan, Y . Pan, G. Baldan, and O. Beijbom, “nuscenes: A multimodal dataset for autonomous driving,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 11 621–11 631

    2020

  24. [32]

    A deep automotive radar detector using the radelft dataset,

    I. Roldan, A. Palffy, J. F. Kooij, D. M. Gavrila, F. Fioranelli, and A. Yarovoy, “A deep automotive radar detector using the radelft dataset,”IEEE Transactions on Radar Systems, 2024

    2024

  25. [33]

    Radical: A synchronized fmcw radar, depth, imu and rgb camera data dataset with low-level fmcw radar signals,

    T.-Y . Lim, S. A. Markowitz, and M. N. Do, “Radical: A synchronized fmcw radar, depth, imu and rgb camera data dataset with low-level fmcw radar signals,”IEEE Journal of Selected Topics in Signal Processing, vol. 15, no. 4, pp. 941–953, 2021

    2021

  26. [34]

    Radarrgbd a multi- sensor fusion dataset for perception with rgb-d and mmwave radar,

    T. Song, J. Ye, A. Guo, G. He, and B. Yang, “Radarrgbd a multi- sensor fusion dataset for perception with rgb-d and mmwave radar,” arXiv preprint arXiv:2505.15860, 2025

    2025 arXiv

  27. [35]

    Multi- class road user detection with 3+ 1d radar in the view-of-delft dataset,

    A. Palffy, E. Pool, S. Baratam, J. F. Kooij, and D. M. Gavrila, “Multi- class road user detection with 3+ 1d radar in the view-of-delft dataset,” IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 4961–4968, 2022

    2022

  28. [36]

    Coloradar: The direct 3d millimeter wave radar dataset,

    A. Kramer, K. Harlow, C. Williams, and C. Heckman, “Coloradar: The direct 3d millimeter wave radar dataset,”The International Journal of Robotics Research, vol. 41, no. 4, pp. 351–360, 2022

    2022

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.