- Rating
- 2
- Confidence
- 3
- Informativeness
- 1
- Presentation
- 1
- Soundness
- 2
Read review
Paper summary
DiffOps combines pretrained diffusion priors with numerical solvers selected according to the measurement operator and noise model for image restoration and scientific imaging. At each noise level, it generates a clean-image reference, holds it fixed during measurement correction initialized from that reference, and then adds noise for the next level. The framework supports spectral solvers, conjugate gradients, Gauss–Newton, ensemble methods, and Poisson EM, with staged DR→GD correction for phase retrieval. Experiments on FFHQ and four scientific imaging tasks demonstrate its applicability and practical value.
Critical strengths
- The framework addresses the limitations of uniform measurement correction by selecting solvers based on operator structure and noise characteristics, making effective use of established numerical methods.
- The paper covers diverse inverse problems and supports the rationale for operator-aware adaptation through mechanism analyses, baseline comparisons, and solver ablations.
- The supplementary material includes parameter settings and implementation code, facilitating reproduction and further extension of the framework.
Critical weaknesses
- The presentation and layout require further refinement. (1) Sections 2 and 3 could distinguish existing methods, approximation assumptions, and new designs more clearly. (2) Some configurations are inconsistent: raw MRI uses CG×2 in Table S7 but CG×20 inner iterations in Table S13. (3) ''Principled probabilistic imaging using diffusion models as plug-and-play priors'' is cited twice in the references. (4) On page 12, the DDNM first author’s name is misspelled as "Yinhui Wang" instead of "Yinhuai Wang." (5) Some figures and tables also have excessive surrounding whitespace and insufficient explanatory detail.
- The practical gains from solver selection have not been sufficiently isolated. Most deterministic solvers achieve similar reconstruction quality on linear FFHQ tasks. A comparison against a single-solver baseline, such as gradient descent across all tasks, using the same diffusion prior, sampling schedule, tuning budget, and hardware would better establish the independent quality or efficiency gains from operator-aware adaptation.
- Some image panels lack visual polish, limiting the overall aesthetic quality of the presentation.
- The methodological novelty of the proposed approach needs to be further clarified. Its main contribution lies in integrating established solvers into diffusion reconstruction based on operator characteristics, with task-specific solver selection and parameter tuning.
Questions for authors
- How is the aggregate quality score in Figure 3 defined? PSNR and SSIM have different preference directions from LPIPS and FID, and their numerical scales differ. Directly averaging them and subtracting a standard deviation therefore does not ensure that higher values indicate better reconstruction. Please specify the metrics, direction adjustments, normalization, and standard-deviation definition for each task.
- For new inverse problems, solver selection and parameter configuration still rely on users’ expertise in numerical optimization, operator structure, and noise models, raising the barrier to use and limiting transferability to new tasks. Could the authors provide explicit solver-selection criteria and parameter-setting guidelines based on operator structure, available derivative information, and noise models?
- The applicability of the framework to latent diffusion models remains unverified. Unlike the latent extension demonstrated by DAPS [a], incorporating a decoder into DiffOps generally makes the composite operator nonlinear, even when the original measurement operator is linear. How should solver selection be adapted, and can the original quality and efficiency advantages be retained?
[a] Improving diffusion inverse problem solving with decoupled noise annealing. CVPR 2025.
4 comments
Q2 (guidelines). The selection rule is stated in §2 (L137–160) and §3.2 (L235–250), with the solver families in Table S1 (Supp. L068) and the per-task choices in Table S13 (Supp. L929); in short:
| Operator structure | Solver |
|---|---|
| Linear, FFT-diagonalizable normal operator | Wiener (closed form) |
| Linear, matrix-free and | CG |
| Smooth nonlinear, usable Jacobian | Gauss–Newton or L-BFGS |
| Zero or unavailable Jacobian | EKI |
| Poisson counts, nonnegative operator | EM |
| Nonconvex magnitude data | DR, then GD |
In the submitted minimal-tuning configuration, is 0 for noiseless data and otherwise shared per noise model, with across Gaussian FFHQ operators; inner iterations are fixed per solver family (Table S13). Sharing is a default choice motivated by relative insensitivity to moderate variations, not a restriction. The additional operator-specific selection of and the solver on 100 validation images assesses sensitivity and potential gains from more extensive tuning; its solver choices agree with the table: CG for SR and inpainting and Wiener for motion blur.
Q3 (latent priors). All experiments use pixel-space priors as in the FFHQ and InverseBench protocols for all baselines. With a latent prior the composite operator is nonlinear even for linear . That places it in DiffOps's nonlinear branch: Gauss–Newton or L-BFGS with the decoder Jacobian, or EKI if the decoder is treated as a black box. Alternatively, the correction can run as in pixel space and be re-encoded and decoded, as in ReSample. However, we also additionally validate DiffOps-Latent (GN) under the same evaluation protocol and operators used in our other image-restoration tasks on FFHQ256, 100 validation images. At each noise level, we approximately solve
We use Levenberg–Marquardt initialized at the denoised latent estimate . Each step uses 20 conjugate-gradient iterations to approximately solve
where is the Levenberg–Marquardt damping parameter. We compare 50 and 100 NFE with LatentDAPS at 1K/4K NFE. Entries are PSNR↑ / SSIM↑ / LPIPS↓; the best value for each metric within each operator is bold.
| Operator | LatentDAPS (1K/4K NFE) | L-GN, 50 NFE | L-GN, 100 NFE |
|---|---|---|---|
| Gaussian blur | 27.93 / 0.764 / 0.234 | 28.98 / 0.821 / 0.200 | 29.09 / 0.825 / 0.195 |
| 4× SR | 27.48 / 0.801 / 0.182 | 29.10 / 0.832 / 0.185 | 29.17 / 0.834 / 0.179 |
| Motion blur | 27.00 / 0.814 / 0.283 | 28.98 / 0.820 / 0.194 | 29.12 / 0.828 / 0.188 |
| Box inpainting | 23.99 / 0.802 / 0.194 | 24.06 / 0.823 / 0.172 | 24.53 / 0.841 / 0.147 |
| Nonlinear blur | 28.11 / 0.713 / 0.235 | 27.87 / 0.810 / 0.207 | 28.85 / 0.830 / 0.185 |
| HDR | 25.94 / 0.751 / 0.223 | 24.82 / 0.831 / 0.191 | 25.22 / 0.835 / 0.187 |
At 100 NFE, DiffOps-Latent improves SSIM and LPIPS mostly. These results support DiffOps's applicability to latent diffusion priors and show that its correction framework can improve latent-space reconstruction with substantially fewer diffusion-network evaluations.
We add this to the conclusion as future validation, but comprehensively evaluating latent priors is beyond this submission.
I thank the authors for their detailed responses and for addressing some of my concerns. However, the manuscript still requires further improvement, and I am therefore inclined to maintain my current score at this stage.
Thank you for reviewing our responses; we have further improved our manuscript and uploaded the revised PDF.
We thank the reviewer for the careful reading and the concrete presentation comments. Please also see the general response.
W1 (presentation). All items are fixed in the revision:
Table S13 now lists CG×2 for raw MRI (Supp. L939); Tables S7/S8 (Supp. L687, L714) and the released preset use 2.
The duplicate Wu et al. entry is merged (L645–650).
"Yinhuai Wang" is corrected (L638).
Figures now span the full line width with less surrounding whitespace, captions give more detail, and the image panels are polished.
The paper states what DiffOps shares with prior work and where it differs (L116–117, L126–127, L194–196). In short:
Shared form. The correction objective has the same decoupled form, with an anchor at the denoiser output, as in DAPS, DiffPIR and related methods.
Different role. In those methods the anchor approximates the diffusion prior and keeps the correction consistent with the diffusion trajectory. In DiffOps it is a geometric local constraint around the denoiser output (L194–196), which turns each noise level into a standalone classical problem. That is why classical solver selection applies.
What follows. The diffusion signal-to-noise ratio enters only through λt=β/σt2, which sets the weight, step size and conditioning of whichever solver is used. Solvers can therefore be combined across noise levels to match the stages of the diffusion, e.g., DR at high noise and GD at low noise for phase retrieval.
W2 (single-solver control). The paper contains this control for all seven FFHQ operators (Table S4, Supp. L541; Table S5, Supp. L559; Table 2, L408–419), with the same prior, schedule and hardware. On linear operators the solvers tie, as the analysis predicts for a strictly convex quadratic. On HDR, nonlinear blur and phase retrieval, GD loses by 1.6 dB, 2.2 dB and 22 points of success.
The same control on the scientific tasks, with A10 runtimes:
Raw MRI ties at two iterations, where correction depth, not the solver, is the limit.
W4 (novelty). Our core contribution is the framework outlined in W1: the geometric anchor reduces each noise level to a standalone classical subproblem under a fully decoupled formulation, allowing the correction mechanism to be tailored to the operator structure and adapted across noise levels. W2 demonstrates the resulting gains: our matched solver outperforms baselines when the forward operator violates default correction assumptions, matches them otherwise, and improves overall sampling efficiency. Whereas existing methods such as DDS, ReSample, and DPS adhere strictly to the prior manifold via a single, rigid correction mechanism, DiffOps readily accommodates operator classes none of them evaluate—including zero-Jacobian clipping, network blur, closure quantities, exact Poisson likelihoods, and staged non-convex corrections. Furthermore, our formulation naturally extends to physics-based forward models and neural operators by treating data correction as a general non-linear inverse problem.
Q1 (Figure 3). Each panel's vertical axis is the mean normalized reconstruction-quality score: an equal-weight average of that task's quality metrics, without standard-deviation subtraction. Each metric is first direction-corrected so that higher is better (LPIPS, FID and relative L2 error are reversed), then min–max normalized across the methods in that panel. Heights are comparable only within a panel. Metrics: FFHQ PSNR, SSIM, LPIPS, FID; MRI and microscopy PSNR, SSIM; inverse scattering PSNR, SSIM, relative L2; black hole shift-aligned and 20 µas beam-blurred PSNR. The revision states this definition in the main text. The original text mentioned one standard deviation because we initially intended to include a standard-deviation indication in the visualization. However, the final figure displays the mean scores, and the corresponding text was inadvertently left unchanged. We will remove the outdated reference to one standard deviation and explicitly state the correct definition in the Figure 3 caption (L340).