Pith. sign in

REVIEW 5 major objections 5 minor 1 cited by

FLRNet: A Deep Learning Method for Regressive Reconstruction of Flow Field From Limited Sensor Measurements

T0 review · 5 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read FLRNet reconstructs full flow fields from sparse sensors with lower error than POD and MLP baselines.

desk verdict A plausible VAE-based flow reconstruction architecture worth a referee, but the reported gains are single stochastic draws with no error bars and the closest baseline is missing. read the letter →

arxiv 2411.13815 v1 pith:PTGRAFNM submitted 2024-11-21 physics.flu-dyn cs.LG

Phong C. H. Nguyen , Joseph B. Choi , Quang-Trung Luu This is my paper · ORCID
classification physics.flu-dyncs.LG
keywords flowfieldreconstructionsparsesensorsvariationalautoencoderFourierfeaturesperceptuallossspectralbiascylinderwakeregressive
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 introduces FLRNet, a regressive deep-learning method for reconstructing full velocity fields from a handful of point sensors. The authors argue that the measurement map from field to sensors is ill-conditioned and non-invertible, so instead of learning the inverse directly they first learn a low-dimensional latent representation of the flow field with a variational autoencoder, then train a dense network to map sensor readings to that latent space. FLRNet adds Fourier feature layers and a perceptual loss during autoencoder training to counter spectral bias, the tendency of neural networks to smooth out high-frequency flow structures. Tested on flow around a circular cylinder at Reynolds numbers 300 to 1000, with 8 to 32 sensors and three sensor layouts, FLRNet reports lower mean absolute error than MLP and POD baselines in every tested configuration, and smaller error growth when sensor noise is added. If these results hold, the method is a reusable, one-time-trained alternative to optimization-based reconstruction for unsteady wake flows.

What carries the argument

The central mechanism is a two-stage deep-learning pipeline. Stage one trains a Fourier-feature-based variational autoencoder, where the encoder receives both the flow snapshot and pixel coordinate fields, applies a Gaussian Fourier mapping $\gamma(x) = [\cos(Bx_1), \sin(Bx_1), \cos(Bx_2), \sin(Bx_2)]$, and learns a latent distribution; the decoder reconstructs the field from a sampled latent vector, with training guided by the standard VAE loss plus a perceptual loss computed from a pretrained image-network's feature maps. Stage two freezes the autoencoder and trains a five-layer MLP to map sensor readings to latent mean and variance vectors, then samples and decodes to reconstruct the field. The Fourier features counter spectral bias by controlling frequency falloff, while the perceptual loss adds feature-space fidelity beyond per-pixel error, and the two-stage design keeps the expensive representation learning separate from the sensor-to-latent regression.

What would settle it

Retrain every model ten times with different random seeds on the same data and compare error distributions: if the best baseline's best seed beats FLRNet's worst seed, or if the interquartile ranges overlap the reported gaps, the consistent-outperformance conclusion fails.

Watch

Extended reading notes

Core claim

The central claim is that flow-field reconstruction from sparse sensors improves by separating the problem into two learned stages: a variational autoencoder that compresses full flow fields into a low-dimensional latent space, and a fully connected network that maps sensor measurements to that latent space. FLRNet's autoencoder is augmented with Fourier feature layers and a perceptual loss, which the paper argues mitigates spectral bias and preserves high-frequency wake structures that blurry POD and MLP reconstructions lose. Quantitatively, the paper reports FLRNet with Fourier features reaching a mean absolute error of 0.016 m/s at 32 randomly placed sensors, compared with 0.036 m/s for the next best baseline, POD, and 0.046 m/s for MLP (Table 1). The same ranking holds across sensor counts of 8, 16, and 32, across three sensor layouts, across Reynolds numbers from 350 to 1000, and under increasing levels of additive Gaussian sensor noise, supporting the paper's conclusion that FLRNet is both the most accurate and the most robust method tested.

Load-bearing premise

The reported error margins come from single training runs with no error bars, so the claim that FLRNet consistently outperforms every baseline rests on those single-point differences being larger than run-to-run variation.

Editorial extensions

If this is right

  • For the tested cylinder-wake regime, a single trained FLRNet model can replace per-case optimization or repeated training, reconstructing unseen Reynolds numbers inside and at the boundaries of the training range with the lowest error among compared methods.
  • The internal comparison shows Fourier features alone outperform the perceptual-loss variant, suggesting that correcting frequency bias is the primary driver of FLRNet's accuracy gain over the direct-learning baselines.
  • Reconstruction accuracy improves monotonically with sensor count and is best for random sensor layouts, but FLRNet keeps the lowest error even for the hardest layout tested, sensors clustered around the cylinder.
  • Under additive Gaussian sensor noise up to a standard deviation of 0.5 m/s, FLRNet's mean absolute error grows more slowly than that of MLP and POD, supporting deployment with realistic noisy sensors.
  • Because the decoder is frozen during sensor-mapping training, the mapping network alone could be retrained for a new sensor configuration, leaving the learned flow representation intact.

Reading between the lines

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

  • The autoencoder's latent space is learned from clean full fields and could potentially be reused for other inverse problems, such as reconstructing from different sensor modalities or from partial domain observations, without retraining the decoder; the paper only hints at this flexibility.
  • The perceptual loss relies on a pretrained image network, and the paper does not test whether a flow-specific feature extractor would improve or harm transfer; comparing these choices is a direct experimental extension.
  • The reported margins are single-run estimates, so the natural next benchmark is to train each model multiple times with different random seeds and compare error distributions; if the interquartile ranges overlap the reported gaps, the 'consistently outperforms' claim would need qualification.
  • If Fourier features are truly countering spectral bias, the same architectural fix should reduce blur in other field-reconstruction tasks with sharp fronts, such as multiphase interfaces or shock-containing flows, which is a testable prediction outside the cylinder-wake setting.
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

5 major / 5 minor

Summary. The manuscript introduces FLRNet, a two-stage deep learning method for reconstructing flow fields from sparse sensor measurements. In stage one, a variational autoencoder (VAE) with Fourier feature layers and an additional perceptual loss is trained to learn a low-dimensional latent representation of high-fidelity flow snapshots. In stage two, a fully connected network is trained to map sensor measurements to this latent representation; at inference, the sensor-derived latent variable is passed through the frozen decoder to reconstruct the full field. The method is validated on a cylinder-wake benchmark with Reynolds numbers between 300 and 1000, varying the number of sensors (8, 16, 32), three sensor layouts, noise levels, and test Reynolds numbers. The abstract and Section 4 claim that FLRNet consistently outperforms MLP and POD baselines in all tested scenarios and is the most robust to noise.

Significance. If the claims were supported, the contribution would be valuable: the paper proposes a plausible remedy for spectral bias in field reconstruction by combining Fourier feature mapping and perceptual loss within a VAE, and it addresses the practical problem of sparse-sensor reconstruction across multiple sensor configurations and flow conditions. The experimental design covers several practically relevant variations (sensor count, layout, noise, Reynolds number). However, the current evidence is not sufficient to establish the central claim. Key weaknesses are the absence of uncertainty quantification for a stochastic method, the omission of the most relevant VAE baseline (Dubois et al. [6]), ambiguous definitions of the two FLRNet variants, and an overstatement of the generalization claim given that all test Reynolds numbers lie within the training range. These issues are load-bearing for the headline 'consistently outperformed' claim, so the manuscript requires major revision.

major comments (5)
  1. [Section 3.3] The inference procedure is stochastic: after computing the predicted mean and variance, the latent variable is randomly sampled and passed to the decoder. Consequently, every reconstructed field and every MAE in Table 1 and Figures 4-8 is a random variable for a fixed checkpoint. The paper reports only point estimates, with no standard deviations, confidence intervals, or number of latent samples averaged per test snapshot. Without such variability measures, the reported margins (e.g., Table 1: FLRNet Fourier feat. 0.016 vs. POD 0.036 at 32 sensors) cannot be distinguished from favorable stochastic draws. I request repeated inference runs and/or multiple training seeds with mean and spread reported.
  2. [Section 4.1 and Table 1] The most relevant baseline is missing. The paper cites Dubois et al. [6], which introduced VAE-based reconstruction from limited measurements and reported that a deep variational autoencoder achieves the highest accuracy and robustness among the methods tested. Since FLRNet is also VAE-based, a direct comparison against the Dubois VAE (adapted to the same cylinder-wake data) is necessary to support the claim that FLRNet 'consistently outperformed other baselines.' Comparing only against MLP and POD, whose poorer performance the paper itself attributes to known limitations, does not establish advantage over the state of the art.
  3. [Section 1 (Main contributions) and Section 4] There is an internal inconsistency between the claimed scope and the presented experiments. The introduction states that FLRNet was 'trained and tested using two benchmark problems with various flow conditions,' but Section 4 presents only one benchmark: flow around a circular obstacle. No second benchmark appears anywhere in the results. This contradiction must be resolved, either by removing the claim of two benchmark problems or by actually including the second benchmark in the validation.
  4. [Table 1 and Section 3.2] The definition of the two FLRNet variants is ambiguous. The method description in Section 3.2 presents Fourier features and perceptual loss as two components of the single proposed architecture. Table 1 then lists 'FLRNet (Percep. loss)' and 'FLRNet (Fourier feat.)' as two variants, but the text never states whether these are ablations (one component included, the other excluded) or whether each variant includes both components with one emphasized. The hyperparameters for each variant (e.g., m, sigma, latent dimension, perceptual loss weight) are also not specified per variant. This ambiguity undermines the conclusion drawn in Section 4.1 that Fourier features are more effective than perceptual loss against spectral bias.
  5. [Section 4.5 and Figure 8] The generalization claim is over-stated relative to the evidence. The paper states that FLRNet shows 'generalizability across different flow conditions' and discusses 'interpolation vs. extrapolation' at the training boundaries, but the test Reynolds numbers shown in Figure 8 (350, 550, 750, 1000) all lie within the stated training range of 300 to 1000. No test case outside this range is reported, so the method's behavior in extrapolation is not actually demonstrated. In addition, the manuscript does not specify which Reynolds numbers are used in the training and test splits, making it impossible to assess the difficulty of the interpolation task. Please provide the exact Re values in the train/test split and, if the claim is about generalization, include hold-out cases outside the training range.
minor comments (5)
  1. [Abstract and Introduction] The text contains typos and grammatical errors, including 'an variational autoencoder' in the abstract, 'different Reynold numbers' instead of 'Reynolds numbers,' and 'the a snapshot' in Section 3.1. These should be corrected.
  2. [Equation (7)] In Equation (7), the text says 'where u is the reconstructed flow field, ˆu is its corresponding ground truth,' which reverses the notation used elsewhere in the paper (u is the ground truth and ˆu is the reconstruction). Please fix the variable names in the definition.
  3. [Section 4.2] The term 'vortex shredding' appears in the discussion of Figure 5; the correct fluid-dynamics term is 'vortex shedding.'
  4. [Section 3.2] The perceptual loss uses Inception-V3 pretrained on ImageNet, but the manuscript does not describe how the scalar flow-field snapshots are converted into the three-channel, network-compatible input format (e.g., normalization, resizing, channel composition). This detail is needed for reproducibility.
  5. [General] The paper does not state whether code or data will be made available. Given the reproducibility concerns raised above, a data/code availability statement would be important for a computational paper of this type.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FLRNet is validated by an external train/test comparison, not by a self-referential derivation.

full rationale

The paper's central claim is empirical: FLRNet reconstructs cylinder-wake velocity fields from sparse sensors and is compared against POD and MLP baselines on held-out test cases (Table 1, Figs. 3-8). No equation defines the predicted field in terms of a fitted parameter, and no fitted parameter is renamed as a prediction. The VAE latent representation, sensor-to-latent mapping G_z (Eq. 3), and decoder reconstruction (Eq. 4) form a standard supervised pipeline whose evaluation uses external baselines and test cases not used for training. Self-citations (refs. 14, 19, 20) appear in related-work and spectral-bias context, but none is load-bearing: the performance comparison does not depend on those cited results. The stochastic latent sampling at inference (Section 3.3) and the lack of error bars are statistical-reporting concerns, not circularity, because the reconstruction is still evaluated against ground-truth fields rather than being constructed to match them. Therefore no circular step can be identified and quoted, and the appropriate finding is no significant circularity.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The method rests on several hand-set hyperparameters and on domain assumptions about simulation fidelity, sensor model, and the transferability of ImageNet-trained features to flow fields. No new physical entities are introduced.

free parameters (6)
  • Number of Fourier features m = 4
    Set to 4 in Section 3.2; chosen by hand, not optimized, and affects the frequency content of the position encoding.
  • Gaussian variance sigma for B = 5
    Set to 5 in Section 3.2; controls the frequency falloff of Fourier features, hand-picked.
  • Latent dimension = not reported
    The size of z is never stated; it is a key capacity knob for the VAE and the sensor-to-latent mapping.
  • Perceptual loss weight = not reported
    Equation (7) adds L_perceptual to the VAE loss, but the weighting coefficient is never specified.
  • Learning rate = 1e-5
    Adam optimizer learning rate reported in Section 3.4; a hand-set hyperparameter.
  • MLP architecture (5 layers x 128 neurons) = 128 hidden units per layer
    Architecture choice in Section 3.3; no justification or search reported.
assumptions (5)
  • domain assumption The flow is governed by incompressible Navier-Stokes equations and the finite volume simulation provides ground truth fields.
    Section 4 states the numerical solution is obtained by solving Navier-Stokes with finite volume method; the entire training and test data rely on this simulation being accurate.
  • domain assumption Pointwise sensor measurements y = H(u) are exactly the velocity magnitude at the sensor locations.
    Section 3.1 defines H as providing punctual measurement; the paper assumes this operator is known and that sensors measure the same quantity as the training target.
  • ad hoc to paper A pretrained Inception-V3 (on ImageNet) provides a meaningful perceptual feature space for comparing flow field images.
    Section 3.2 applies ImageNet-trained Inception-V3 to velocity magnitude fields without discussing the domain shift; if this transfer is invalid, the perceptual loss term may not behave as claimed.
  • domain assumption Fourier feature mapping with random B ~ N(0, sigma^2) mitigates spectral bias in the convolutional VAE.
    Section 3.2 cites Rahimi and Recht and Tancik et al.; the paper assumes this technique carries over to flow fields and to a convolutional (rather than MLP) architecture.
  • domain assumption The latent space of the VAE is smooth and continuous enough for a small MLP to map sensor measurements to the correct latent code.
    Section 3.1 and 3.3 assume the VAE latent space is well-distributed; without this, the mapping network Gz cannot work, but no analysis of latent space structure is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FLRNet: A Deep Learning Method for Regressive Reconstruction of Flow Field From Limited Sensor Measurements." pith.science (2026). https://pith.science/paper/PTGRAFNM

@misc{pith2026241113815,
  author       = {Pith},
  title        = {Pith review of: FLRNet: A Deep Learning Method for Regressive Reconstruction of Flow Field From Limited Sensor Measurements},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PTGRAFNM}},
  note         = {Machine review of arXiv:2411.13815}
}
read the original abstract

Many applications in computational and experimental fluid mechanics require effective methods for reconstructing the flow fields from limited sensor data. However, this task remains a significant challenge because the measurement operator, which provides the punctual sensor measurement for a given state of the flow field, is often ill-conditioned and non-invertible. This issue impedes the feasibility of identifying the forward map, theoretically the inverse of the measurement operator, for field reconstruction purposes. While data-driven methods are available, their generalizability across different flow conditions (\textit{e.g.,} different Reynold numbers) remains questioned. Moreover, they frequently face the problem of spectral bias, which leads to smooth and blurry reconstructed fields, thereby decreasing the accuracy of reconstruction. We introduce FLRNet, a deep learning method for flow field reconstruction from sparse sensor measurements. FLRNet employs an variational autoencoder with Fourier feature layers and incorporates an extra perceptual loss term during training to learn a rich, low-dimensional latent representation of the flow field. The learned latent representation is then correlated to the sensor measurement using a fully connected (dense) network. We validated the reconstruction capability and the generalizability of FLRNet under various fluid flow conditions and sensor configurations, including different sensor counts and sensor layouts. Numerical experiments show that in all tested scenarios, FLRNet consistently outperformed other baselines, delivering the most accurate reconstructed flow field and being the most robust to noise.

Figures

Figures reproduced from arXiv: 2411.13815 by the authors.

Figure 1
Figure 1. The overall architecture design. latent representation of the corresponding flow field. For the VAE, we use a Fourier-feature-based, fully convolutional architecture with an additional loss term besides the conventional VAE loss, namely the perceptual loss. These added extra features to the neural network design and training will enable the VAE to learn perceptually rich features that are aware of the dynamic charac… view at source ↗
Figure 2
Figure 2. Numerical experiment setting for the flow around cylindrical test problem. (a) Dimension of the examined [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Reconstruction result of FLRNet and other baselines at different times of the simulation. FLRNet reconstructed [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Error profile analysis. We computed the average MAE across the whole test dataset at six different horizontal [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Analysis of the reconstruction error w.r.t the temporal evolution of the flow field. The reconstruction error [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Effect of different sensor configurations on the performance of FLRNet and other baselines. (a) Effect of [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Effect of noise in sensor measurement. Compared to other baselines, FLRNet is the most robust method as its [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Reconstruction error of FLRNet compared to other baselines for different flow conditions. [PITH_FULL_IMAGE:figures/full_fig_p011_8.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. A Physics-Augmented GraphGPS Framework for the Reconstruction of 3D Riemann Problems from Sparse Data

    physics.flu-dyn 2025-05 conditional novelty 6.0 of 10

    A graph transformer with shock-aware attention and guided message-passing reconstructs 3D Riemann problem flows from 10% observed points more accurately than the benchmark ML models tested.

Reference graph

Works this paper leans on

31 extracted references · 24 canonical work pages · cited by 1 Pith paper

  1. [6]

    Machine learning for fluid flow recon- struction from limited measurements

    Pierre Dubois, Thomas Gomez, Laurent Planckaert, and Laurent Perret. Machine learning for fluid flow recon- struction from limited measurements. Journal of Computational Physics, 448:110733, 2022

    work page 2022

  2. [1]

    Callaham, Kazuki Maeda, and Steven L

    Jared L. Callaham, Kazuki Maeda, and Steven L. Brunton. Robust flow reconstruction from limited measurements via sparse representation. Phys. Rev. Fluids, 4:103907, Oct 2019. 12

    work page 2019

  3. [2]

    Rowley and Scott T.M

    Clarence W. Rowley and Scott T.M. Dawson. Model reduction for flow analysis and control. Annual Review of Fluid Mechanics, 49:387–417, 2017

    work page 2017

  4. [3]

    Deep-learning assisted reduced order model for high-dimensional flow prediction from sparse data

    Jiaxin Wu, Dunhui Xiao, and Min Luo. Deep-learning assisted reduced order model for high-dimensional flow prediction from sparse data. Physics of Fluids, 35(10), 2023

    work page 2023

  5. [4]

    Caverly and J

    D. Caverly and J. Nedi´c. Estimating wind velocity and direction using sparse sensors on a cylinder. Physics of Fluids, 34(9), 2022

    work page 2022

  6. [5]

    Flow field reconstruction of 2d hypersonic inlets based on a variational autoencoder

    Zuwei Tan, Runze Li, and Yufei Zhang. Flow field reconstruction of 2d hypersonic inlets based on a variational autoencoder. Aerospace, 10(9), 2023

    work page 2023

  7. [7]

    Noack, and Steven L

    Jean-Christophe Loiseau, Bernd R. Noack, and Steven L. Brunton. Sparse reduced-order modelling: sensor-based dynamics to full-state estimation. Journal of Fluid Mechanics, 844:459–490, 2018

    work page 2018

  8. [8]

    Mons, J.-C

    V . Mons, J.-C. Chassaing, T. Gomez, and P. Sagaut. Reconstruction of unsteady viscous flows using data assimilation schemes. Journal of Computational Physics, 316:255–280, 2016

    work page 2016

Show all 31 references
  1. [9]

    Benjamin Erichson, Lionel Mathelin, Zhewei Yao, Steven L

    N. Benjamin Erichson, Lionel Mathelin, Zhewei Yao, Steven L. Brunton, Michael W. Mahoney, and J. Nathan Kutz. Shallow neural networks for fluid flow reconstruction with limited sensors. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences , 476(2...

    2020

  2. [10]

    Deep learning reconstruction of high-reynolds-number turbulent flow field around a cylinder based on limited sensors

    Rui Li, Baiyang Song, Yaoran Chen, Xiaowei Jin, Dai Zhou, Zhaolong Han, Wen-Li Chen, and Yong Cao. Deep learning reconstruction of high-reynolds-number turbulent flow field around a cylinder based on limited sensors. Ocean Engineering, 304:117857, 2024

    2024

  3. [11]

    Recfno: A resolution- invariant flow and heat field reconstruction method from sparse observations via fourier neural operator

    Xiaoyu Zhao, Xiaoqian Chen, Zhiqiang Gong, Weien Zhou, Wen Yao, and Yunyang Zhang. Recfno: A resolution- invariant flow and heat field reconstruction method from sparse observations via fourier neural operator. Interna- tional Journal of Thermal Sciences , 195:108619, 2024

    2024

  4. [12]

    A physics-informed deep learning model to reconstruct turbulent wake from random sparse data

    Peixing Xie, Rui Li, Yaoran Chen, Baiyang Song, Wen-Li Chen, Dai Zhou, and Yong Cao. A physics-informed deep learning model to reconstruct turbulent wake from random sparse data. Physics of Fluids, 36(6):065145, 06 2024

    2024

  5. [13]

    On the eigenvector bias of fourier feature networks: From regression to solving multi-scale pdes with physics-informed neural networks

    Sifan Wang, Hanwen Wang, and Paris Perdikaris. On the eigenvector bias of fourier feature networks: From regression to solving multi-scale pdes with physics-informed neural networks. Computer Methods in Applied Mechanics and Engineering, 384:113938, 2021

    2021

  6. [14]

    Phong C. H. Nguyen, Xinlun Cheng, Shahab Azarfar, Pradeep Kumar Seshadri, Yen Thi Nguyen, Munho Kim, Sanghun Choi, H. S. Udaykumar, and Stephen Baek. Parcv2: Physics-aware recurrent convolutional neural networks for spatiotemporal dynamics modeling. ArXiv, abs/2402.12503, 2024

    2024 arXiv

  7. [15]

    Brunton, Scott T.M

    Kunihiko Taira, Steven L. Brunton, Scott T.M. Dawson, Clarence W. Rowley, Tim Colonius, Beverley J. McKeon, Oliver T. Schmidt, Stanislav Gordeyev, Vassilios Theofilis, and Lawrence S. Ukeiley. Modal analysis of fluid flows: An overview. AIAA Journal, 55(12):4013 – 4041, 2017

    2017

  8. [16]

    Callaham, Kazuki Maeda, and Steven L

    Jared L. Callaham, Kazuki Maeda, and Steven L. Brunton. Robust flow reconstruction from limited measurements via sparse representation. Phys. Rev. Fluids, 4:103907, Oct 2019

    2019

  9. [17]

    Noack, Konstantin Afanasiev, Marek Morzy ´nski, Gilead Tadmor, and Frank Thiele

    Bernd R. Noack, Konstantin Afanasiev, Marek Morzy ´nski, Gilead Tadmor, and Frank Thiele. A hierarchy of low-dimensional models for the transient and post-transient cylinder wake. Journal of Fluid Mechanics, 497:335 – 363, 2003

    2003

  10. [18]

    Brunton, Joshua L

    Steven L. Brunton, Joshua L. Proctor, and J. Nathan Kutz. Discovering governing equations from data by sparse identification of nonlinear dynamical systems. Proceedings of the National Academy of Sciences , 113(15):3932– 3937, 2016

    2016

  11. [19]

    Nguyen, Pradeep K

    Xinlun Cheng, Phong C.H. Nguyen, Pradeep K. Seshadri, Mayank Verma, Zoë J. Gray, Jack T. Beerman, H.S. Udaykumar, and Stephen S. Baek. Physics-aware recurrent convolutional neural networks for modeling multiphase compressible flows. International Journal of Multiphase Flow , 1...

    2024

  12. [20]

    Phong C. H. Nguyen, Yen-Thi Nguyen, Joseph B. Choi, Pradeep K. Seshadri, H. S. Udaykumar, and Stephen S. Baek. Parc: Physics-aware recurrent convolutional neural networks to assimilate meso scale reactive mechanics of energetic materials. Science Advances, 9(17):eadd6868, 2023

    2023

  13. [21]

    Arnault, J

    A. Arnault, J. Dandois, and J.-M. Foucaut. Comparison of stochastic estimation methods with conditional events optimization for the reconstruction of the flow around a supercritical airfoil in transonic conditions. Computers & Fluids, 136:436–455, 2016. 13

    2016

  14. [22]

    A hybrid deep learning framework for unsteady periodic flow field reconstruction based on frequency and residual learning

    Xingwen Peng, Xingchen Li, Xiaoqian Chen, Xianqi Chen, and Wen Yao. A hybrid deep learning framework for unsteady periodic flow field reconstruction based on frequency and residual learning. Aerospace Science and Technology, 141:108539, 2023

    2023

  15. [23]

    Supervised learning method for the physical field reconstruction in a nanofluid heat transfer problem

    Tianyuan Liu, Yunzhu Li, Qi Jing, Yonghui Xie, and Di Zhang. Supervised learning method for the physical field reconstruction in a nanofluid heat transfer problem. International Journal of Heat and Mass Transfer , 165:120684, 2021

    2021

  16. [24]

    A deep learning method based on partition modeling for reconstructing temperature field

    Xingwen Peng, Xingchen Li, Zhiqiang Gong, Xiaoyu Zhao, and Wen Yao. A deep learning method based on partition modeling for reconstructing temperature field. International Journal of Thermal Sciences , 182:107802, 2022

    2022

  17. [25]

    Deep Learning

    Ian Goodfellow, Yoshua Bengio, and Aaron Courville. Deep Learning. MIT Press, 2016

    2016

  18. [26]

    Kingma and Max Welling

    Diederik P. Kingma and Max Welling. An introduction to variational autoencoders. CoRR, abs/1906.02691, 2019

    1906 arXiv

  19. [27]

    Random features for large-scale kernel machines

    Ali Rahimi and Benjamin Recht. Random features for large-scale kernel machines. In J. Platt, D. Koller, Y . Singer, and S. Roweis, editors, Advances in Neural Information Processing Systems , volume 20. Curran Associates, Inc., 2007

    2007

  20. [28]

    Srinivasan, Ben Mildenhall, Sara Fridovich-Keil, Nithin Raghavan, Utkarsh Singhal, Ravi Ramamoorthi, Jonathan T

    Matthew Tancik, Pratul P. Srinivasan, Ben Mildenhall, Sara Fridovich-Keil, Nithin Raghavan, Utkarsh Singhal, Ravi Ramamoorthi, Jonathan T. Barron, and Ren Ng. Fourier features let networks learn high frequency functions in low dimensional domains. NeurIPS, 2020

    2020

  21. [29]

    Perceptual losses for real-time style transfer and super-resolution

    Justin Johnson, Alexandre Alahi, and Li Fei-Fei. Perceptual losses for real-time style transfer and super-resolution. In Bastian Leibe, Jiri Matas, Nicu Sebe, and Max Welling, editors,Computer Vision – ECCV 2016, pages 694–711, Cham, 2016. Springer International Publishing

    2016

  22. [30]

    Rethinking the inception architecture for computer vision

    Christian Szegedy, Vincent Vanhoucke, Sergey Ioffe, Jonathon Shlens, and Zbigniew Wojna. Rethinking the inception architecture for computer vision. CoRR, abs/1512.00567, 2015

    2015 arXiv

  23. [31]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition , pages 248–255, 2009. 14

    2009

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.