Pith. sign in

REVIEW 4 major objections 5 minor 79 references

Generalized Task-Driven Medical Image Quality Enhancement with Gradient Promotion

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read GradProm adds a recognition model's gradient to medical image enhancement training only when the two gradients point the same way, preventing the recognition task from biasing the enhancer.

desk verdict Solid empirical story undermined by an ill-typed update rule and an invalid proof; salvageable but not as written. read the letter →

arxiv 2501.01114 v1 pith:5NDV4QXI submitted 2025-01-02 cs.CV

Dong Zhang , Kwang-Ting Cheng This is my paper · ORCID
classification cs.CV
keywords imagequalityenhancementmedicalprocessingtask-auxiliarylearningmulti-taskgradientpromotioncosinesimilaritydenoisingsuper-resolution
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 proposes GradProm, a training strategy for task-driven medical image quality enhancement in which an image enhancement model and a visual recognition model are trained together, but the recognition model's gradient is used to update the enhancement model only when the two gradients point in the same direction, as measured by cosine similarity. The paper's central claim is that this gating prevents the auxiliary recognition task from biasing the enhancement model's optimization direction, while still allowing recognition information to help when it is aligned with the enhancement objective. The authors prove that with a sufficiently small learning rate, the GradProm update keeps a non-negative inner product with the enhancement gradient and can reach a local minimum of the enhancement loss. Across four medical datasets (skin lesion, COVID CT, histopathology, and ultrasound), they report consistent gains in PSNR and SSIM over joint training, frozen-parameter training, and several recent denoising and super-resolution baselines, together with improved diagnosis and segmentation accuracy. If the claim is right, medical images can be restored to be both more human-perceivable and more machine-usable without changing the network architecture or adding any inference cost.

What carries the argument

The mechanism is the gradient gate $\max(0, \cos(G_{IP}, G_{VR}))$, a scalar computed from the cosine similarity between the enhancement-model gradient and the recognition-model gradient. At every training step the enhancement update is $G_{IP} + G_{VR}\max(0,\cos(G_{IP},G_{VR}))$, so the auxiliary gradient is switched on when the two gradients agree and switched off when they disagree. This gate is the object the theoretical analysis works with: the non-negativity of the inner product between the total update and $G_{IP}$ is what the local-minimum proof relies on.

What would settle it

Take any task-driven IQE network in which the enhancement and recognition sub-models have disjoint parameter sets and compute $\cos(\nabla_\theta L_{IP}, \nabla_\phi L_{VR})$; because the two gradients live in different vector spaces, the cosine similarity is undefined, so Lemma 3.1's update rule cannot be executed as written unless the paper explicitly redefines $G_{VR}$ as $\nabla_\theta L_{VR}$.

Watch

Extended reading notes

Core claim

GradProm partitions a task-driven IQE system into a mainstream image enhancement model with parameters $\theta$ and an auxiliary visual recognition model with parameters $\phi$, and updates only the enhancement model's parameters. With the paper's notation $G_{IP}=\nabla_\theta L_{IP}(\theta)$ and $G_{VR}=\nabla_\phi L_{VR}(\phi)$, GradProm's update for the enhancement parameters is $\theta^{t+1} = \theta^t - \alpha^t\big(G_{IP}^t + G_{VR}^t \max(0, \cos(G_{IP}^t, G_{VR}^t))\big)$; when the cosine similarity is negative, the auxiliary gradient term is dropped. The paper claims that this rule guarantees the inner product of the total update with the enhancement gradient $\nabla_\theta L_{IP}$ is non-negative, so the enhancement model's descent direction is never biased toward the recognition model's direction, and that GradProm can achieve a local minimum of the mainstream model. Empirically, the authors report state-of-the-art PSNR/SSIM for denoising and super-resolution on ISIC 2018, COVID-CT, Lizard, and CAMUS, with better downstream accuracy and mIoU than the compared training strategies, and show that the method generalizes across datasets and image modalities without extra data or architecture changes.

Load-bearing premise

The whole method depends on comparing and adding two gradients that are defined with respect to different models' parameters; if those parameter sets do not overlap, the cosine similarity and the update are not well defined.

Editorial extensions

If this is right

  • Task-driven medical image enhancement can be improved without extra data, new architectures, or added inference time; the only extra training-time cost is computing one cosine similarity per update.
  • The gating rule should transfer across medical modalities and task combinations, since the paper shows gains on skin lesion, CT, histology, and ultrasound data with both classification and segmentation as the auxiliary task.
  • If the proof holds, GradProm converges to a local minimum of the enhancement model even with a recognition model attached, so the recognition task can assist enhancement without pulling it toward the recognition model's own optimum.
  • Cross-dataset generalization improves: training on one medical dataset and testing on another yields higher PSNR/SSIM with GradProm than with the frozen-parameter baseline.
  • Combining two auxiliary recognition tasks can hurt rather than help, so the method is best used with auxiliary tasks whose feature requirements are similar to each other and to the enhancement task.

Reading between the lines

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

  • A natural extension is to replace the hard on/off threshold at $\cos=0$ with a continuous weight, for example scaling $G_{VR}$ by a smooth function of the cosine similarity; the paper explicitly declines to explore intermediate alignments, so whether finer gating helps is an open empirical question.
  • The proof only establishes a non-negative inner product with the enhancement gradient, which is a much weaker condition than a descent inequality on a nonconvex network; a stronger convergence guarantee would require bounding how much the auxiliary term can increase $L_{IP}$.
  • Part of the gain might come from the gate acting as a denoiser of the enhancement gradient itself, rather than from the recognition task's semantic content; a control experiment replacing $G_{VR}$ with random vectors whose alignment distribution matches the real one would isolate that effect.
  • The same gating logic applies to any cascade of a main model and an auxiliary model, such as detection or compression branches, because the rule only needs two gradient vectors and a cosine similarity.
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 / 5 minor

Summary. The paper proposes GradProm, a training strategy for task-driven medical image quality enhancement (IQE) in which the enhancement model (IP) and a downstream visual recognition model (VR) are trained with a gradient gate: when the gradients of the two sub-models have non-negative cosine similarity, the update to the IP parameters is augmented by the VR gradient; otherwise only the IP gradient is used. The authors claim a theoretical guarantee that this update converges to a local minimum of the IP loss without being biased by the VR loss, and they report state-of-the-art denoising and super-resolution results on ISIC 2018, COVID-CT, Lizard, and CAMUS, together with downstream diagnosis and segmentation performance.

Significance. If the theoretical claim and the algorithm were correctly stated, the paper would offer a simple, architecture-agnostic training modification with potentially broad applicability to medical image enhancement, and the experimental study is extensive: four public datasets, multiple noise levels and degradation types, two auxiliary tasks, supervised/unsupervised settings, domain generalization, and comparisons with several recent baselines. The paper also reports repeated runs with standard deviations, which is a strength. However, the central formalism is ill-typed, the proof of Lemma 3.1 contains a false logical inference, and several claimed improvement values in the text do not match the tables. These are not presentation issues: they undermine the theoretical contribution and the reproducibility of the method as written.

major comments (4)
  1. [§3.2, Eq. (5)-(6) and Lemma 3.1] The method is not well-posed as written. The paper defines G_IP = ∇_θ L_IP(θ) and G_VR = ∇_ϕ L_VR(ϕ), with θ and ϕ parameters of IP and VR respectively, and Eq. (5) writes L_total(θ_T) = L_IP(θ) + λ L_VR(ϕ), indicating that θ and ϕ are different parameter blocks. The cosine similarity s = cos(G_IP, G_VR) and the update rule in Lemma 3.1, θ^{t+1}_T := θ^t_T − α_t(G^t_IP + G^t_VR max(0, cos(G^t_IP, G^t_VR))), require G_VR to be a tangent vector in the same parameter space as G_IP. As defined, G_VR is not; it is a gradient with respect to a disjoint parameter vector. If one pads the two gradients to the full parameter space, their cosine similarity is zero because their nonzero entries live in disjoint coordinates, which would make the method identical to training IP alone and contradict the reported gains. The paper never states that G_VR is intended to mean ∇_θ L_VR, the gradient of the VR loss with respect to the IP parameters through the composite function VR(IP(X)); the sentence in §3.2 that says "we can use both L_IP(θ) and L_VR(ϕ) to update G_IP of the mainstream IP, i.e., G_IP = ∇_θ(L_IP(θ)+L_VR(ϕ))" reassigns G_IP inconsistently. This ill-typed definition makes the central algorithm and the lemma it supports unreproducible from the text.
  2. [§3.3, Lemma 3.1 proof, Eq. (9)] The proof contains a false inference. After expanding the inner product, the paper considers ⟨G_VR max(0, cos(G_IP, G_VR)), ∇L_IP⟩ = 0 and then states: "Since max(0, cos(·)) ≥ 0, we can infer ... either ⟨G_VR, ∇L_IP⟩ = 0 or ⟨∇L_IP, G_VR⟩ = 0. In either case, we can conclude that ∇L_IP = 0, since the cosine similarity between two non-zero vectors is positive." This is invalid: two nonzero vectors can be orthogonal, and even if G_VR were zero, the conclusion ∇L_IP = 0 would not follow. Consequently, the claimed implication that a zero inner product forces the IP gradient to vanish is false, and the central assertion that GradProm "can achieve the local minimum of the mainstream IP model" is not established by the given argument.
  3. [§3.3, convergence claim] The proof's overall strategy is insufficient for the claimed conclusion. Showing ⟨Δθ^t_T, ∇L_IP^t⟩ ≥ 0 only establishes that the update direction is not an ascent direction at the current point; it does not imply convergence to a local minimum. A local-minimum guarantee would require a sufficient-descent condition relative to the gradient norm and a suitable learning-rate schedule, not merely the phrase "under the condition that α_t is as small as possible." Without such a condition, the theorem is at best a statement about non-increase of a linearized objective. This is a load-bearing gap because the paper's advertised contribution includes the theoretical guarantee of unbiased convergence.
  4. [§4.4, A-1 paragraph] Several claimed improvement magnitudes in the text do not match the reported tables. In the A-1 paragraph, after correctly reporting gains over Frozen-params for Table 1, the text says "GradProm improves ESTR by 5.544/0.082, 7.085/0.108, 4.534/0.240, and 5.898/0.197 PSNR/SSIM" for Table 2. These differences do not correspond to GradProm versus Frozen-params in Table 2; for example, at σ=0.05 the frozen-params unsupervised PSNR is 32.826 and GradProm is 33.525, a difference of 0.699, while 5.544 equals 36.825 − 31.281, i.e., the difference between GradProm (supervised) and Benchmark-i. This makes the strength of the reported improvements unclear and needs correction.
minor comments (5)
  1. [§3.2 and Figure 3] Figure 3 and the surrounding text are inconsistent about whether VR parameters are updated: the Abstract says GradProm "updates only parameters of the image enhancement model," but §3.2 says "L_VR(ϕ) of the auxiliary V_R is only used to updated by its own gradient G_VR" and Eq. (6) explicitly includes ∇_ϕ L_VR in the total gradient. Please clarify whether VR parameters are frozen, updated, or jointly trained.
  2. [§3.2, equation after Eq. (6)] The notation in the two-case display for G_T is inconsistent with the definitions: the first case writes ∇_θ(L_IP(θ)+λ L_VR(ϕ)) for the IP block, but earlier G_IP was defined as ∇_θ L_IP(θ), so the left-hand side should be a total update direction, not G_IP. Please introduce separate symbols for the update direction and the individual gradients.
  3. [§4.1, Experiment Setup] The paper states that λ is set to 10^-4 but provides no sensitivity analysis or ablation over λ; since λ controls the contribution of the VR loss, a small study or at least a discussion of its role would strengthen the empirical claims.
  4. [§4.4, Table 2 vs Table 3] The text in A-5 says the performance gain of GradProm in the challenging scenario is reduced compared to Tables 1 and 2, but Table 3 shows GradProm outperforming Frozen-params by much larger relative SSIM margins (e.g., 0.088 and 0.144) than many entries in Tables 1 and 2; this qualitative statement should be checked against the actual numbers.
  5. [Throughout] There are several typographical errors and unclear phrases, including "the the image enhancement model" (Abstract), "mainstream mainstream" (§4.4 qualitative results), "Experimrntal" (Table 4 caption), and duplicated SSIM header in Table 7. These do not affect the technical content but should be corrected.

Circularity Check

1 steps flagged · score 2.0 of 10

Low-severity self-definitional theorem: the claimed 'unbiased' guarantee restates the max(0, cos) gate; the empirical benchmark claims are not circular.

  1. self definitional [Section 3.3, Lemma 3.1 and its proof (Eqs. 7-9); update rule also given in Section 3.2.]
    "θ^{t+1}_T := θ^t_T − α_t (G^t_IP + G^t_VR max(0, cos(G^t_IP, G^t_VR))) ... ⟨G^t_VRmax(0, cos(G^t_IP, G^t_VR)), ∇L^t_IP⟩ ≥ 0, since the cosine similarity between G^t_IP and G^t_VRis always non-negative."

    The claimed guarantee that the IP update is not biased by the VR gradient is inserted by the definition of the update rule. The auxiliary gradient appears only multiplied by max(0, cos(G_IP, G_VR)), so the proof's 'observation 2' simply restates that the gating coefficient is non-negative on the branch that uses G_VR. The non-negative inner product in Eq. 7 is therefore true by construction, not by any property of the losses or networks. The further inference meant to convert this into a local-minimum statement—that a vanishing inner product implies ∇L_IP = 0—is not valid, since two non-zero vectors can be orthogonal. Thus the theorem reduces to the gate rather than establishing independent convergence.

full rationale

The paper's empirical claims are self-contained: GradProm is compared with externally defined baselines (SR-ResNet, ESTR, U-SAID, D2SM, WinNet, ADAP, etc.) on four public datasets, and no parameter is fitted to the target metric and then reported as a prediction. References to the authors' prior work (e.g., [3], [31], [53], [64]) are related-work or implementation-style citations and are not load-bearing for the main result. The only closeness to circularity is the theoretical analysis in Section 3.3: Lemma 3.1's non-negative inner product is guaranteed by the algorithm's own gate, because the auxiliary gradient is added only when its cosine with the IP gradient is non-negative. Consequently, the 'not biased by VR' conclusion is a restatement of the update rule rather than an independent theorem. There is also a separate, non-circular rigor issue: G_VR is defined as ∇_φ L_VR(φ) while being added to an update for θ, so the cosine similarity and the additive update are not well-defined if θ and φ are disjoint parameter blocks; this is a correctness concern, not a circularity. Overall, the central empirical contribution remains externally validated, so the circularity score is low.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The central empirical claim rests on few fitted numbers, but the theoretical claim rests on shifting definitions of the gradients and on an invalid optimization principle. No new entities are introduced.

free parameters (1)
  • λ (loss weight for L_VR) = 10^-4
    Set empirically in Section 4.1; no sensitivity analysis is provided, and the reported behavior may depend on this choice.
assumptions (3)
  • domain assumption The cosine similarity between G_IP and G_VR is a meaningful measure of whether the auxiliary model is useful.
    Used throughout Section 3.2 to decide when to include the auxiliary gradient; requires both gradients in the same vector space and assumes alignment implies benefit.
  • ad hoc to paper A non-negative inner product between the update direction and ∇L_IP guarantees convergence to a local minimum for a sufficiently small learning rate.
    Invoked in Lemma 3.1; the proof treats zero inner product as implying ∇L_IP=0, which is false because orthogonality is possible.
  • ad hoc to paper An aligned auxiliary gradient never harms the main task.
    The method assumes positive cosine similarity is always beneficial; no ablation tests the sign of the gate or the magnitude of λ.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Generalized Task-Driven Medical Image Quality Enhancement with Gradient Promotion." pith.science (2026). https://pith.science/paper/5NDV4QXI

@misc{pith2026250101114,
  author       = {Pith},
  title        = {Pith review of: Generalized Task-Driven Medical Image Quality Enhancement with Gradient Promotion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5NDV4QXI}},
  note         = {Machine review of arXiv:2501.01114}
}
read the original abstract

Thanks to the recent achievements in task-driven image quality enhancement (IQE) models like ESTR, the image enhancement model and the visual recognition model can mutually enhance each other's quantitation while producing high-quality processed images that are perceivable by our human vision systems. However, existing task-driven IQE models tend to overlook an underlying fact -- different levels of vision tasks have varying and sometimes conflicting requirements of image features. To address this problem, this paper proposes a generalized gradient promotion (GradProm) training strategy for task-driven IQE of medical images. Specifically, we partition a task-driven IQE system into two sub-models, i.e., a mainstream model for image enhancement and an auxiliary model for visual recognition. During training, GradProm updates only parameters of the image enhancement model using gradients of the visual recognition model and the image enhancement model, but only when gradients of these two sub-models are aligned in the same direction, which is measured by their cosine similarity. In case gradients of these two sub-models are not in the same direction, GradProm only uses the gradient of the image enhancement model to update its parameters. Theoretically, we have proved that the optimization direction of the image enhancement model will not be biased by the auxiliary visual recognition model under the implementation of GradProm. Empirically, extensive experimental results on four public yet challenging medical image datasets demonstrated the superior performance of GradProm over existing state-of-the-art methods.

Figures

Figures reproduced from arXiv: 2501.01114 by the authors.

Figure 1
Figure 1. The evolution of medical image quality enhancement (IQE) in visual recognition. The IQE model is represented by dots of different colors in [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The guided backpropagation visualizations for different vision [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. (a) Joint training puts the upstream and downstream models [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Qualitative result comparisons with the baseline ESTR [ [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Class activation maps (CAMs) result comparisons with the baseline model ESTR [ [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Visualization result comparisons with state-of-the-art methods on ISIC 2018 [ [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Visualization results on the genuine low-quality images. Sam [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

79 extracted references · 79 canonical work pages

  1. [1]

    Exploring simple and transferable recognition-aware image processing,

    Z. Liu, H. Wang, T. Zhou, Z. Shen, B. Kang, E. Shelhamer, and T. Darrell, “Exploring simple and transferable recognition-aware image processing,” IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 2022

    work page 2022

  2. [2]

    Machine learning for medical imaging: methodological failures and recommendations for the future,

    G. Varoquaux and V . Cheplygina, “Machine learning for medical imaging: methodological failures and recommendations for the future,” NPJ Digital Medicine, vol. 5, no. 1, p. 48, 2022

    work page 2022

  3. [3]

    Deep learning for medical image segmentation: tricks, challenges and future directions,

    D. Zhang, Y. Lin, H. Chen, Z. Tian, X. Yang, J. Tang, and K. T. Cheng, “Deep learning for medical image segmentation: tricks, challenges and future directions,” arXiv, 2022

    work page 2022

  4. [4]

    Deep learning in medical image analysis,

    D. Shen, G. Wu, and H.-I. Suk, “Deep learning in medical image analysis,” Annual Review of Biomedical Engineering, vol. 19, pp. 221– 248, 2017

    work page 2017

  5. [5]

    Medical image analysis using convolutional neural networks: a review,

    S. M. Anwar, M. Majid, A. Qayyum, M. Awais, M. Alnowami, and M. K. Khan, “Medical image analysis using convolutional neural networks: a review,” Journal of Medical Systems , vol. 42, pp. 1–13, 2018

    work page 2018

  6. [6]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 770–778. 13

    work page 2016

  7. [7]

    A survey on vision transformer,

    K. Han, Y. Wang, H. Chen, X. Chen, J. Guo, Z. Liu, Y. Tang, A. Xiao, C. Xu, Y. Xu et al., “A survey on vision transformer,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 45, no. 1, pp. 87–110, 2022

    work page 2022

  8. [8]

    Fully convolutional net- works for semantic segmentation,

    J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional net- works for semantic segmentation,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp. 3431–3440

    work page 2015

Show all 79 references
  1. [9]

    Automated pulmonary embolism detection from ctpa images us- ing an end-to-end convolutional neural network,

    Y. Lin, J. Su, X. Wang, X. Li, J. Liu, K.-T. Cheng, and X. Yang, “Automated pulmonary embolism detection from ctpa images us- ing an end-to-end convolutional neural network,” in Medical Image Computing and Computer Assisted Intervention (MICCAI) , 2019, pp. 280–288

    2019

  2. [10]

    Fast and low-gpu-memory abdomen ct organ segmentation: the flare challenge,

    J. Ma, Y. Zhang, S. Gu, X. An, Z. Wang, C. Ge, C. Wang, F. Zhang, Y. Wang, Y. Xu et al. , “Fast and low-gpu-memory abdomen ct organ segmentation: the flare challenge,” Medical Image Analysis , vol. 82, p. 102616, 2022

    2022

  3. [11]

    Future frame prediction for robot-assisted surgery,

    X. Gao, Y. Jin, Z. Zhao, Q. Dou, and P .-A. Heng, “Future frame prediction for robot-assisted surgery,” in Information Processing in Medical Imaging (IPMI), 2021, pp. 533–544

    2021

  4. [12]

    A review of image enhancement techniques in medical imaging,

    A. Jawdekar and M. Dixit, “A review of image enhancement techniques in medical imaging,” Machine Intelligence and Smart Systems (MISS), pp. 25–33, 2021

    2021

  5. [13]

    An image quality enhancement scheme employ- ing adolescent identity search algorithm in the nsst domain for multimodal medical image fusion,

    J. Jose, N. Gautam, M. Tiwari, T. Tiwari, A. Suresh, V . Sundararaj, and M. Rejeesh, “An image quality enhancement scheme employ- ing adolescent identity search algorithm in the nsst domain for multimodal medical image fusion,” Biomedical Signal Processing and Control, vol. 66...

    2021

  6. [14]

    A dictionary learning approach for poisson image deblurring,

    L. Ma, L. Moisan, J. Yu, and T. Zeng, “A dictionary learning approach for poisson image deblurring,” IEEE Transactions on Medical Imaging, vol. 32, no. 7, pp. 1277–1289, 2013

    2013

  7. [15]

    Self- supervised domain adaptation for breaking the limits of low- quality fundus image quality enhancement,

    Q. Hou, P . Cao, J. Wang, X. Liu, J. Yang, and O. R. Zaiane, “Self- supervised domain adaptation for breaking the limits of low- quality fundus image quality enhancement,” arXiv, 2023

    2023

  8. [16]

    Transforming medical imaging with transformers? a comparative review of key properties, current progresses, and future perspec- tives,

    J. Li, J. Chen, Y. Tang, C. Wang, B. A. Landman, and S. K. Zhou, “Transforming medical imaging with transformers? a comparative review of key properties, current progresses, and future perspec- tives,” Medical Image Analysis, p. 102762, 2023

    2023

  9. [17]

    A review of the deep learning methods for medical images super resolution problems,

    Y. Li, B. Sixou, and F. Peyrin, “A review of the deep learning methods for medical images super resolution problems,” IRBM, vol. 42, no. 2, pp. 120–133, 2021

    2021

  10. [18]

    Multimodal multi-head con- volutional attention with various kernel sizes for medical image super-resolution,

    M.-I. Georgescu, R. T. Ionescu, A.-I. Miron, O. Savencu, N.-C. Ristea, N. Verga, and F. S. Khan, “Multimodal multi-head con- volutional attention with various kernel sizes for medical image super-resolution,” in Proceedings of the IEEE/CVF Winter Conference on Applications of ...

    2023

  11. [19]

    Medical image denoising techniques: a review,

    R. Patil and S. Bhosale, “Medical image denoising techniques: a review,” International Journal on Engineering, Science and Technology, vol. 4, no. 1, pp. 21–33, 2022

    2022

  12. [20]

    Image super- resolution using progressive generative adversarial networks for medical image analysis,

    D. Mahapatra, B. Bozorgtabar, and R. Garnavi, “Image super- resolution using progressive generative adversarial networks for medical image analysis,” Computerized Medical Imaging and Graph- ics, vol. 71, pp. 30–39, 2019

    2019

  13. [21]

    Psnr vs ssim: imperceptibility quality assess- ment for image steganography,

    D. R. I. M. Setiadi, “Psnr vs ssim: imperceptibility quality assess- ment for image steganography,” Multimedia Tools and Applications, vol. 80, no. 6, pp. 8423–8444, 2021

    2021

  14. [22]

    Transformers in medical imaging: A survey,

    F. Shamshad, S. Khan, S. W. Zamir, M. H. Khan, M. Hayat, F. S. Khan, and H. Fu, “Transformers in medical imaging: A survey,” arXiv, 2022

    2022

  15. [23]

    Task-driven feature pooling for image classification,

    G.-S. Xie, X.-Y. Zhang, X. Shu, S. Yan, and C.-L. Liu, “Task-driven feature pooling for image classification,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2015, pp. 1179– 1187

    2015

  16. [24]

    Classification-driven dynamic image enhance- ment,

    V . Sharma, A. Diba, D. Neven, M. S. Brown, L. Van Gool, and R. Stiefelhagen, “Classification-driven dynamic image enhance- ment,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 4033–4041

    2018

  17. [25]

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

    J. Johnson, A. Alahi, and L. Fei-Fei, “Perceptual losses for real- time style transfer and super-resolution,” in European Conference on Computer Vision (ECCV), 2016, pp. 694–711

    2016

  18. [26]

    Nima: Neural image assessment,

    H. Talebi and P . Milanfar, “Nima: Neural image assessment,”IEEE Transactions on Image Processing, vol. 27, no. 8, pp. 3998–4011, 2018

    2018

  19. [27]

    Super-resolution to improve classification accuracy of low-resolution images,

    L. Jaffe, S. Sundram, and C. Martinez-Nieves, “Super-resolution to improve classification accuracy of low-resolution images,” Tech. Rep. 19, Stanford University, Tech. Rep., 2017

    2017

  20. [28]

    Defense against adversarial attacks using high-level representation guided denoiser,

    F. Liao, M. Liang, Y. Dong, T. Pang, X. Hu, and J. Zhu, “Defense against adversarial attacks using high-level representation guided denoiser,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 1778–1787

    2018

  21. [29]

    Visual interpretability for deep learning: a survey,

    Q.-s. Zhang and S.-C. Zhu, “Visual interpretability for deep learning: a survey,” Frontiers of Information Technology & Electronic Engineering, vol. 19, no. 1, pp. 27–39, 2018

    2018

  22. [30]

    Interpretability- guided inductive bias for deep learning based medical image,

    D. Mahapatra, A. Poellinger, and M. Reyes, “Interpretability- guided inductive bias for deep learning based medical image,” Medical Image Analysis, vol. 81, p. 102551, 2022

    2022

  23. [31]

    Causal intervention for weakly-supervised semantic segmentation,

    D. Zhang, H. Zhang, J. Tang, X.-S. Hua, and Q. Sun, “Causal intervention for weakly-supervised semantic segmentation,” in Advances in Neural Information Processing Systems (NeurIPS) , 2020, pp. 655–666

    2020

  24. [32]

    Joint learning of intrinsic images and semantic segmentation,

    A. S. Baslamisli, T. T. Groenestege, P . Das, H.-A. Le, S. Karaoglu, and T. Gevers, “Joint learning of intrinsic images and semantic segmentation,” in European Conference on Computer Vision (ECCV) , 2018, pp. 286–302

    2018

  25. [33]

    Joint learning of saliency detection and weakly supervised semantic segmentation,

    Y. Zeng, Y. Zhuge, H. Lu, and L. Zhang, “Joint learning of saliency detection and weakly supervised semantic segmentation,” in Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 7223–7233

    2019

  26. [34]

    Gradient harmonized single-stage de- tector,

    B. Li, Y. Liu, and X. Wang, “Gradient harmonized single-stage de- tector,” inThe Association for the Advancement of Artificial Intelligence (AAAI), 2019, pp. 8577–8584

    2019

  27. [35]

    What makes training multi- modal classification networks hard?

    W. Wang, D. Tran, and M. Feiszli, “What makes training multi- modal classification networks hard?” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020, pp. 12 695–12 705

    2020

  28. [36]

    Auxiliary learning for deep multi-task learning,

    Y. Liu, B. Zhuang, C. Shen, H. Chen, and W. Yin, “Auxiliary learning for deep multi-task learning,” arXiv, 2019

    2019

  29. [37]

    Prompt-aligned gradient for prompt tuning,

    B. Zhu, Y. Niu, Y. Han, Y. Wu, and H. Zhang, “Prompt-aligned gradient for prompt tuning,” arXiv, 2022

    2022

  30. [38]

    Photo- realistic single image super-resolution using a generative adver- sarial network,

    C. Ledig, L. Theis, F. Husz ´ar, J. Caballero, A. Cunningham, A. Acosta, A. Aitken, A. Tejani, J. Totz, Z. Wang et al. , “Photo- realistic single image super-resolution using a generative adver- sarial network,” in Proceedings of the IEEE Conference on Computer Vision and Patt...

    2017

  31. [39]

    U-net: Convolutional networks for biomedical image segmentation,

    O. Ronneberger, P . Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in Medical Image Computing and Computer Assisted Intervention (MICCAI) , 2015, pp. 234–241

    2015

  32. [40]

    Skin lesion analysis toward melanoma detection 2018: A chal- lenge hosted by the international skin imaging collaboration (isic),

    N. Codella, V . Rotemberg, P . Tschandl, M. E. Celebi, S. Dusza, D. Gutman, B. Helba, A. Kalloo, K. Liopyris, M. Marchetti et al., “Skin lesion analysis toward melanoma detection 2018: A chal- lenge hosted by the international skin imaging collaboration (isic),” arXiv, 2019

    2018

  33. [41]

    Covid-ct- dataset: a ct scan dataset about covid-19,

    X. Yang, X. He, J. Zhao, Y. Zhang, S. Zhang, and P . Xie, “Covid-ct- dataset: a ct scan dataset about covid-19,” arXiv, 2020

    2020

  34. [42]

    Lizard: a large-scale dataset for colonic nuclear instance segmentation and classification,

    S. Graham, M. Jahanifar, A. Azam, M. Nimir, Y.-W. Tsang, K. Dodd, E. Hero, H. Sahota, A. Tank, K. Benes et al. , “Lizard: a large-scale dataset for colonic nuclear instance segmentation and classification,” in Proceedings of the IEEE/CVF International Conference on Computer Vi...

    2021

  35. [43]

    Deep learning for segmentation using an open large-scale dataset in 2d echocardiography,

    S. Leclerc, E. Smistad, J. Pedrosa, A. Østvik, F. Cervenansky, F. Espinosa, T. Espeland, E. A. R. Berg, P .-M. Jodoin, T. Grenier et al., “Deep learning for segmentation using an open large-scale dataset in 2d echocardiography,” IEEE Transactions on Medical Imaging, vol. 38, n...

    2019

  36. [44]

    Higcin: Hierarchical graph-based cross inference network for group activity recogni- tion,

    R. Yan, L. Xie, J. Tang, X. Shu, and Q. Tian, “Higcin: Hierarchical graph-based cross inference network for group activity recogni- tion,” IEEE transactions on pattern analysis and machine intelligence , vol. 45, no. 6, pp. 6955–6968, 2020

    2020

  37. [45]

    Generative adversarial network in medical imaging: A review,

    X. Yi, E. Walia, and P . Babyn, “Generative adversarial network in medical imaging: A review,” Medical Image Analysis , vol. 58, p. 101552, 2019

    2019

  38. [46]

    Progressive instance- aware feature learning for compositional action recognition,

    R. Yan, L. Xie, X. Shu, L. Zhang, and J. Tang, “Progressive instance- aware feature learning for compositional action recognition,”IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 45, no. 8, pp. 10 317–10 330, 2023

    2023

  39. [47]

    A new frame- work of designing iterative techniques for image deblurring,

    M. Zhang, G. S. Young, Y. Tie, X. Gu, and X. Xu, “A new frame- work of designing iterative techniques for image deblurring,” Pattern Recognition, vol. 124, p. 108463, 2022

    2022

  40. [48]

    Deep semantic statistics matching (d2sm) denoising network,

    K. Mei, V . M. Patel, and R. Huang, “Deep semantic statistics matching (d2sm) denoising network,” in European Conference on Computer Vision (ECCV), 2022, pp. 384–400

    2022

  41. [49]

    Multitask learning,

    R. Caruana, “Multitask learning,” Machine Learning, vol. 28, no. 1, pp. 41–75, 1997. 14

    1997

  42. [50]

    An overview of multi-task learning,

    Y. Zhang and Q. Yang, “An overview of multi-task learning,” National Science Review, vol. 5, no. 1, pp. 30–43, 2018

    2018

  43. [51]

    Multi-task learning as multi-objective optimization,

    O. Sener and V . Koltun, “Multi-task learning as multi-objective optimization,” in Advances in Neural Information Processing Systems (NeurIPS), 2018, pp. 824–836

    2018

  44. [52]

    Deepsaliency: Multi-task deep neural network model for salient object detection,

    X. Li, L. Zhao, L. Wei, M.-H. Yang, F. Wu, Y. Zhuang, H. Ling, and J. Wang, “Deepsaliency: Multi-task deep neural network model for salient object detection,” IEEE Transactions on Image Processing, vol. 25, no. 8, pp. 3919–3930, 2016

    2016

  45. [53]

    Cae-great: Convolutional-auxiliary efficient graph reasoning transformer for dense image predictions,

    D. Zhang, Y. Lin, J. Tang, and K.-T. Cheng, “Cae-great: Convolutional-auxiliary efficient graph reasoning transformer for dense image predictions,” International Journal of Computer Vision , vol. 132, no. 5, pp. 1502–1520, 2024

    2024

  46. [54]

    Residual multi-task learning for facial landmark localization and expression recognition,

    B. Chen, W. Guan, P . Li, N. Ikeda, K. Hirasawa, and H. Lu, “Residual multi-task learning for facial landmark localization and expression recognition,” Pattern Recognition , vol. 115, p. 107893, 2021

    2021

  47. [55]

    From show to tell: a survey on deep learning- based image captioning,

    M. Stefanini, M. Cornia, L. Baraldi, S. Cascianelli, G. Fiameni, and R. Cucchiara, “From show to tell: a survey on deep learning- based image captioning,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 1, pp. 539–559, 2022

    2022

  48. [56]

    Auxiliary tasks in multi-task learning,

    L. Liebel and M. K ¨orner, “Auxiliary tasks in multi-task learning,” arXiv, 2018

    2018

  49. [57]

    Gradient surgery for multi-task learning,

    T. Yu, S. Kumar, A. Gupta, S. Levine, K. Hausman, and C. Finn, “Gradient surgery for multi-task learning,” in Advances in Neural Information Processing Systems (NeurIPS), 2020, pp. 5824–5836

    2020

  50. [58]

    On the effect of auxiliary tasks on representation dynamics,

    C. Lyle, M. Rowland, G. Ostrovski, and W. Dabney, “On the effect of auxiliary tasks on representation dynamics,” in International Conference on Artificial Intelligence and Statistics (ICAIS) . PMLR, 2021, pp. 1–9

    2021

  51. [59]

    Auxiliary task reweighting for minimum-data learning,

    B. Shi, J. Hoffman, K. Saenko, T. Darrell, and H. Xu, “Auxiliary task reweighting for minimum-data learning,” in Advances in Neural Information Processing Systems (NeurIPS), 2020, pp. 7148–7160

    2020

  52. [60]

    Self-supervised generalisation with meta auxiliary learning,

    S. Liu, A. Davison, and E. Johns, “Self-supervised generalisation with meta auxiliary learning,” in Advances in Neural Information Processing Systems (NeurIPS), 2019, pp. 148–160

    2019

  53. [61]

    Adaptive auxiliary task weighting for reinforcement learning,

    X. Lin, H. Baweja, G. Kantor, and D. Held, “Adaptive auxiliary task weighting for reinforcement learning,” in Advances in Neural Information Processing Systems (NeurIPS), 2019, pp. 148–160

    2019

  54. [62]

    Adapting auxiliary losses using gradient similarity,

    Y. Du, W. M. Czarnecki, S. M. Jayakumar, M. Farajtabar, R. Pas- canu, and B. Lakshminarayanan, “Adapting auxiliary losses using gradient similarity,” arXiv, 2018

    2018

  55. [63]

    Im- agenet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Im- agenet: A large-scale hierarchical image database,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2009, pp. 248–255

    2009

  56. [64]

    Self- regulation for semantic segmentation,

    D. Zhang, H. Zhang, J. Tang, X.-S. Hua, and Q. Sun, “Self- regulation for semantic segmentation,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2021, pp. 6953– 6963

    2021

  57. [65]

    Atten- tion deeplabv3+: Multi-level context attention mechanism for skin lesion segmentation,

    R. Azad, M. Asadi Aghbolaghi, M. Fathy, and S. Escalera, “Atten- tion deeplabv3+: Multi-level context attention mechanism for skin lesion segmentation,” in Proceedings of the European Conference on Computer Vision (ECCV), 2020, pp. 251–266

    2020

  58. [66]

    A novel fuzzy hierarchical fusion attention convolution neural network for med- ical image super-resolution reconstruction,

    C. Wang, X. Lv, M. Shao, Y. Qian, and Y. Zhang, “A novel fuzzy hierarchical fusion attention convolution neural network for med- ical image super-resolution reconstruction,” Information Sciences , vol. 622, pp. 424–436, 2023

    2023

  59. [67]

    Unet++: A nested u-net architecture for medical image seg- mentation,

    Z. Zhou, M. M. Rahman Siddiquee, N. Tajbakhsh, and J. Liang, “Unet++: A nested u-net architecture for medical image seg- mentation,” in Medical Image Computing and Computer Assisted Intervention (MICCAI), 2018, pp. 3–11

    2018

  60. [68]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga et al. , “Pytorch: An imperative style, high-performance deep learning library,” in Advances in Neural Information Processing Systems (NeurIPS) , 2019, pp. 7148–7160

    2019

  61. [69]

    Adam: A method for stochastic optimiza- tion,

    D. P . Kingma and J. Ba, “Adam: A method for stochastic optimiza- tion,” arXiv, 2014

    2014

  62. [70]

    Winnet: Wavelet-inspired invertible network for image denoising,

    J.-J. Huang and P . L. Dragotti, “Winnet: Wavelet-inspired invertible network for image denoising,” IEEE Transactions on Image Process- ing, vol. 31, pp. 4377–4392, 2022

    2022

  63. [71]

    Single image super-resolution based on direc- tional variance attention network,

    P . Behjati, P . Rodriguez, C. Fern ´andez, I. Hupont, A. Mehri, and J. Gonz `alez, “Single image super-resolution based on direc- tional variance attention network,” Pattern Recognition, vol. 133, p. 108997, 2023

    2023

  64. [72]

    Federated domain gen- eralization for image recognition via cross-client style transfer,

    J. Chen, M. Jiang, Q. Dou, and Q. Chen, “Federated domain gen- eralization for image recognition via cross-client style transfer,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2023, pp. 361–370

    2023

  65. [73]

    Segmentation- aware image denoising without knowing true segmentation,

    S. Wang, B. Wen, J. Wu, D. Tao, and Z. Wang, “Segmentation- aware image denoising without knowing true segmentation,” arXiv, 2019

    2019

  66. [74]

    Deep image denoising with adaptive priors,

    B. Jiang, Y. Lu, J. Wang, G. Lu, and D. Zhang, “Deep image denoising with adaptive priors,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 8, pp. 5124–5136, 2022

    2022

  67. [75]

    Nafssr: stereo image super-resolution using nafnet,

    X. Chu, L. Chen, and W. Yu, “Nafssr: stereo image super-resolution using nafnet,” in Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), 2022, pp. 1239–1248

    2022

  68. [76]

    Residual local feature network for efficient super-resolution,

    F. Kong, M. Li, S. Liu, D. Liu, J. He, Y. Bai, F. Chen, and L. Fu, “Residual local feature network for efficient super-resolution,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 766–776

    2022

  69. [77]

    Dehazing ultrasound using diffusion models,

    T. S. Stevens, F. C. Meral, J. Yu, I. Z. Apostolakis, J.-L. Robert, and R. J. van Sloun, “Dehazing ultrasound using diffusion models,” arXiv preprint arXiv:2307.11204, 2023

    2023 arXiv

  70. [78]

    Semi- supervised segmentation of echocardiography videos via noise- resilient spatiotemporal semantic calibration and fusion,

    H. Wu, J. Liu, F. Xiao, Z. Wen, L. Cheng, and J. Qin, “Semi- supervised segmentation of echocardiography videos via noise- resilient spatiotemporal semantic calibration and fusion,” Medical Image Analysis, vol. 78, p. 102397, 2022

    2022

  71. [79]

    Supervision by denoising,

    S. I. Young, A. V . Dalca, E. Ferrante, P . Golland, C. A. Metzler, B. Fischl, and J. E. Iglesias, “Supervision by denoising,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023

    2023

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.