LensNet: An End-to-End Learning Framework for Empirical Point Spread Function Modeling and Lensless Imaging Reconstruction

Chi-Man Pun; Jiesong Bai; Xiaofeng Zhang; Xuhang Chen; Yihang Dong; YuHao Yin

REVIEW 3 major objections 5 minor 2 cited by

LensNet: An End-to-End Learning Framework for Empirical Point Spread Function Modeling and Lensless Imaging Reconstruction

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read LensNet claims that an end-to-end network with a learnable Coded Mask Simulator and Wiener fusion blocks can reconstruct lensless camera images without fixed PSF calibration, reporting PSNR of 27.46 dB on DiffuserCam and 33.22 dB on MWDNs.

desk verdict Solid empirical lensless-reconstruction paper with a genuine architecture, but the SOTA claim is not yet established because the baseline protocol is missing and the PSF mechanism is under-specified. read the letter →

arxiv 2505.01755 v1 pith:CXCFX4EQ submitted 2025-05-03 eess.IV cs.CV

Jiesong Bai , Yuhao Yin , Yihang Dong , Xiaofeng Zhang , Chi-Man Pun , Xuhang Chen This is my paper · ORCID

classification eess.IVcs.CV

keywords lenslessimagingpointspreadfunctionCodedMaskSimulatorWienerfilteringdeeplearningreconstructionDiffuserCamMWDNsfrequency-domaindeconvolution

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

LensNet claims that lensless image reconstruction improves when the point spread function (PSF) is learned from data rather than fixed by calibration. The paper proposes an end-to-end network that combines spatial-domain compression with frequency-domain Wiener deconvolution, using a Coded Mask Simulator to generate a data-driven PSF estimate during training. On the DiffuserCam and MWDNs benchmarks, LensNet reports PSNR of 27.46 dB and 33.22 dB respectively, beating the previous best learned method on both datasets, along with lower LPIPS values. If true, this means compact lens-free cameras could adapt to noise, misalignment, and changing scenes without explicit PSF calibration.

What carries the argument

The load-bearing component is the Coded Mask Simulator (CMS), a channel-attention module that pools the measurement feature map globally, passes it through a $1\times1$ convolution with a linear activation, and outputs per-channel weights that are treated as the learned intensity distribution of the coded mask and hence of the PSF. That output feeds the Wiener Fusion Block (WNFB), which takes the fast Fourier transform of both the spatial features and the CMS output, applies the Wiener transfer function $H(u,v) = \overline{\mathrm{PSF}}(u,v)/(|\mathrm{PSF}(u,v)|^2 + \delta)$, and returns the restored signal to the spatial domain via the inverse FFT. The surrounding encoder-decoder structure, built from Reconstruction Blocks and Spatial Amplification Modules, fuses these spatial and frequency representations at multiple scales. The paper's argument is that this design lets the network adapt the PSF to the actual imaging conditions during training and deconvolve scattering and diffraction artifacts without hand-calibrated kernels.

What would settle it

Take the learned distribution map from a trained LensNet, treat it as the PSF, and use Eq. (1) to synthesize a measurement from a known ground-truth image; then compare that synthetic measurement with the actual sensor measurement. If the synthesized measurements clearly differ, or if replacing the learned PSF with the measured or calibrated PSF changes reconstruction quality little, then the CMS is not modeling the true PSF and the claimed mechanism is not what drives the reported gains.

Watch

Extended reading notes

Core claim

The central claim, stated on the paper's own terms, is that jointly estimating the optical encoding and inverting it in one network yields reconstructions that outperform both classical solvers and prior learning-based pipelines. LensNet treats the measurement as the convolution $I_{\mathrm{measurement}} = I_{\mathrm{object}} \ast \mathrm{PSF} + \mathrm{noise}$, encodes the measurement with a Spatial Compression Module, and uses a Coded Mask Simulator to produce a learned distribution map intended to represent the coded mask's PSF. A Wiener Fusion Block then applies the transfer function $H(u,v) = \overline{\mathrm{PSF}}(u,v)/(|\mathrm{PSF}(u,v)|^2 + \delta)$ in the Fourier domain to restore the signal at multiple scales. With a combined MSE, SSIM, and LPIPS loss, the trained model reaches PSNR/SSIM/LPIPS of 27.46/0.863/0.099 on DiffuserCam and 33.22/0.960/0.024 on MWDNs, which the paper reports as consistently better than all compared methods. The authors interpret this as evidence that a learnable PSF, rather than a fixed or sparsely calibrated kernel, is the key to high-quality lensless reconstruction.

Load-bearing premise

The entire gain rests on the assumption that the Coded Mask Simulator's channel-attention output is the physical point spread function used in the Wiener deconvolution step; Section 3.3 gives no explicit formula converting those attention weights or the distribution map into $\mathrm{PSF}(x,y)$, so the improvement could come from the learned image-to-image mapping rather than from accurate PSF estimation.

Editorial extensions

If this is right

On DiffuserCam, LensNet reports PSNR 27.46 dB and SSIM 0.863, beating the prior best learned method MWDN (25.74 dB, 0.816) and lowering LPIPS from 0.132 to 0.099.
On MWDNs, LensNet reports PSNR 33.22 dB and LPIPS 0.024, again ahead of MWDN (31.74 dB, 0.030) and of all classical and learning-based baselines in the comparison.
A camera built on this approach would not need per-device PSF calibration or handcrafted preprocessing, because the network learns the PSF from data during training and can adapt to noise and system imperfections.
Reconstructions preserve high-frequency structure better than prior methods, as shown by qualitative comparisons that keep fur, hair, and fine textures sharper than FISTA, ADMM, FlatNet, UDN, and MWDN outputs.
The user study reported in the appendix gives LensNet the highest average ratings across reconstruction accuracy, detail restoration, artifact presence, noise handling, and visual realism among the methods tested.

Reading between the lines

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

Because Section 3.3 does not define how attention weights become $\mathrm{PSF}(x,y)$, the CMS output may functionally be an adaptive frequency filter rather than a faithful optical model; comparing the learned map with a measured PSF would settle which interpretation is correct.
If the physical PSF interpretation fails, LensNet can still be understood as an unrolled Wiener-deconvolution network with a learned regularizer, a weaker but still useful contribution for lensless imaging.
The same joint PSF-estimation-and-inversion idea could be tested on phase-mask cameras, where the PSF depends on wavelength; a per-channel CMS would be a natural extension.
The largest reported gains are on DiffuserCam, so a stress test under simulated PSF mismatch (for example, a slight rotation or shift of the mask) would show whether the learned PSF genuinely adapts or merely overfits the training distribution.

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.

Desk Editor's Note

read the letter

Quick take: LensNet is a competent empirical deep-learning paper for lensless reconstruction. The architecture—encoder-decoder with a learnable Coded Mask Simulator, a Wiener fusion block, and multi-scale spatial amplification—is a genuine combination I haven't seen in prior work, and the reported numbers on DiffuserCam and MWDN are impressive. The ablation in Table 2 shows the components matter, which is the right kind of evidence. If the numbers are real and reproducible, this is a useful contribution to the field.

That said, the paper as written does not establish the SOTA claim, and two gaps need attention. First, Table 1 gives no baseline protocol. The paper doesn't say whether the competing methods were retrained under the same split and resolution or whether the numbers come from previous publications. DiffuserCam has been used with different splits in the literature, and LensNet's loss includes an LPIPS term, while baselines may have been trained with MSE or something else. That makes the PSNR/LPIPS comparisons potentially biased in LensNet's favor. There are also no error bars or significance tests, so the small margins on MWDN (e.g., SSIM 0.960 vs. 0.957) could be noise. The stress-test note is right that this is the weakest link.

Second, the CMS-to-PSF step is under-specified. Section 3.3 describes a channel-attention mechanism that 'implicitly recovers' the PSF, but it never states how the attention output becomes the PSF(u,v) used in Eq. (4). If the CMS output is just a feature map with no explicit role in the Wiener filter, then the physics-based framing is misleading, and the method reduces to a regular image-to-image network. That doesn't kill the empirical contribution, but it does require a correction in claims.

Minor stuff: the ablation is only on MWDN, the user study has no error bars, and training details (learning rate, epochs) are vague. The math itself is standard and clean; I don't see any errors in Eqs. (1)-(5).

Who this is for: anyone working on lensless imaging or computational imaging who wants a strong new baseline. It deserves a serious referee—the architecture is plausible, the experiments are extensive, and the weaknesses are fixable. I'd send it to peer review with major revision, requiring the authors to specify the baseline protocol, run multiple seeds with error bars, and either formalize the CMS-to-PSF connection or soften the 'physics-based' claim.

Referee Report

3 major / 5 minor

Summary. LensNet is an end-to-end deep learning framework for lensless image reconstruction. It combines an encoder-decoder architecture with a learnable Coded Mask Simulator (CMS) intended to estimate the point spread function from data, and Wiener Fusion Blocks that perform frequency-domain deconvolution. The model is trained with a weighted sum of MSE, SSIM, and LPIPS losses. Experiments on the DiffuserCam and MWDNs datasets compare LensNet with classical and learning-based baselines and report PSNR, SSIM, and LPIPS, together with an ablation study, a user study, and a model-complexity analysis.

Significance. If the reported results are reproducible under the claimed protocol, LensNet would offer a practically useful combination of learned PSF modeling and Wiener filtering for lensless imaging. The paper's strengths include a released code link, a broad set of comparisons, an explicit multi-term loss, and an internal ablation study. However, the central state-of-the-art claim rests on a single table whose baseline protocol is unspecified and whose numerical results are reported without measures of variance. In addition, the manuscript does not specify how the CMS output becomes the PSF used in the Wiener filter, which prevents verification of the physics-based modeling claim. These issues are addressable but currently block full confidence in the headline result.

major comments (3)

[§4.4, Table 1] The paper does not state whether the learning-based baselines (TikNet, FlatNet, LenslessGAN, UDN, MWDN) were retrained under the same protocol or whether their numbers were copied from prior publications. Because Eq. (8) trains LensNet with SSIM and LPIPS terms in addition to MSE, and Table 1 reports exactly those three metrics, the comparison is only meaningful if all baselines were trained with the same loss function, data split, image resolution, and preprocessing. Please specify the training protocol for every baseline, including the loss used, the train/test split, and the number of training epochs.
[§3.3, Eq. (4)] The CMS is described as outputting channel attention weights s_c ∈ R^{N×C′×1×1} after global average pooling and a 1×1 convolution, but the paper never explains how this vector-valued attention output is converted into the two-dimensional PSF(x,y) that enters the Wiener transfer function H(u,v) in Eq. (4). Without an explicit formula or an architectural description connecting the CMS output to PSF(x,y) and to the feature map B(u,v) being filtered, the claim that the CMS performs learnable PSF estimation cannot be verified or reproduced.
[§4.3, Table 1; §4.5, Table 2] No error bars, confidence intervals, or statistical significance tests are reported for any quantitative result. Table 1's improvements over the strongest baseline (e.g., PSNR 27.46 vs. 25.74 on DiffuserCam) may be real, but single-run numbers do not establish that LensNet 'consistently outperforms' all competing methods. The same issue applies to the ablation results in Table 2 and the user-study ratings in Table 3, where no measure of inter-participant variability or significance is provided.

minor comments (5)

[§4.5, Table 2] The ablation study does not state which dataset it was performed on, and the variants 'ThreeDown' and 'w original PSF' are not defined precisely. Please specify the dataset, the training protocol, and exactly what 'original PSF' means in the 'w original PSF' row.
[§3.3, Eq. (7)] The text says σ is a '(linear) activation function'; if the activation is linear, then the channel attention operation is affine and the description should say so explicitly and name the activation actually used in the experiments.
[Figure 3] Figure 3 contains labels such as '???' and 'Random Matrix' that are not explained in the caption or in Section 3.2. Please clarify the notation and the role of the random matrix and learnable parameters in the diagram.
[§4.1] The DiffuserCam dataset split of 24,000 training and 1,000 test images should be justified with respect to the splits used by the cited baselines, since different papers in the literature use different train/test partitions for this dataset.
[Abstract and Conclusion] The abstract and conclusion state that LensNet 'establishes a new standard' and 'outperforms state-of-the-art methods'; these claims are stronger than what a single dataset pair and an unspecified baseline protocol can support. Phrasing such as 'outperforms the compared methods under the reported protocol' would be more appropriate.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: LensNet's held-out reconstruction results are not entailed by construction.

full rationale

The paper's central claim is a supervised reconstruction model evaluated on held-out test images from DiffuserCam and MWDNs, so the reported PSNR, SSIM, and LPIPS values are not forced by the training objective or by any equation that reduces to its own input. The learnable Coded Mask Simulator is trained end-to-end from paired measurements and ground-truth images; it is a fitted internal component, not a separately 'predicted' quantity, so no fitted parameter is renamed as a prediction. The Wiener filter in Eq. (4) is a standard frequency-domain deconvolution applied with the network's internal PSF estimate, and although Sec. 3.3 does not give an explicit formula mapping attention weights to PSF(x,y), that is an interpretability gap rather than a circular derivation. The paper's many self-citations appear only in contextual literature reviews and are not load-bearing for the proposed architecture or results. Concerns about baseline retraining protocols and the LPIPS term in the loss are experimental-rigor issues, not circularity, because the test set is held out and the comparison claim is empirical rather than definitional.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim rests on supervised training rather than first principles. The only physically motivated component, the Wiener filter in Eq (4), assumes a PSF that is never explicitly constructed from the CMS output. Most of the numerical burden is carried by learned weights, so the ledger is dominated by fitted parameters rather than axioms.

free parameters (4)

Network weights = 31.18M parameters
All convolutional and attention parameters, including CMS weights W and b in Eq (7), are learned from paired training data and carry the reconstruction performance.
Wiener regularization delta = unspecified
Eq (4) includes delta in the denominator to stabilize the Wiener filter; the paper never gives its value or how it is set.
Loss weights in Eq (8) = 0.2 and 0.2
Chosen by hand to balance SSIM and LPIPS against MSE; no sensitivity analysis is provided.
Architecture hyperparameters = unspecified
Number of scales, channels per scale, and downsampling depth are not fully specified, though the ablation removes one downsampling layer as ThreeDown.

assumptions (3)

domain assumption The forward model is a space-invariant convolution I_measurement = I_object * PSF + noise (Eq 1).
Standard diffraction-based imaging model from Goodman; assumes a single global PSF, while real lensless systems can have spatially varying PSF and misalignment.
ad hoc to paper Channel attention statistics can represent the coded mask and thus the PSF (Sec 3.3).
The paper asserts that global average pooling and 1x1 convolution produce a distribution map encoding the mask, but provides no derivation or measurement showing this map equals the physical PSF used in Eq (4).
domain assumption Paired training data faithfully sample the real lensless imaging conditions (Sec 4.1).
Performance is measured on the same datasets used to train, so generalization to other optical setups or noise levels is not established.

how reviews work

0 comments

Cite this review

Pith. "Pith review of LensNet: An End-to-End Learning Framework for Empirical Point Spread Function Modeling and Lensless Imaging Reconstruction." pith.science (2026). https://pith.science/paper/CXCFX4EQ

@misc{pith2026250501755,
  author       = {Pith},
  title        = {Pith review of: LensNet: An End-to-End Learning Framework for Empirical Point Spread Function Modeling and Lensless Imaging Reconstruction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CXCFX4EQ}},
  note         = {Machine review of arXiv:2505.01755}
}

read the original abstract

Lensless imaging stands out as a promising alternative to conventional lens-based systems, particularly in scenarios demanding ultracompact form factors and cost-effective architectures. However, such systems are fundamentally governed by the Point Spread Function (PSF), which dictates how a point source contributes to the final captured signal. Traditional lensless techniques often require explicit calibrations and extensive pre-processing, relying on static or approximate PSF models. These rigid strategies can result in limited adaptability to real-world challenges, including noise, system imperfections, and dynamic scene variations, thus impeding high-fidelity reconstruction. In this paper, we propose LensNet, an end-to-end deep learning framework that integrates spatial-domain and frequency-domain representations in a unified pipeline. Central to our approach is a learnable Coded Mask Simulator (CMS) that enables dynamic, data-driven estimation of the PSF during training, effectively mitigating the shortcomings of fixed or sparsely calibrated kernels. By embedding a Wiener filtering component, LensNet refines global structure and restores fine-scale details, thus alleviating the dependency on multiple handcrafted pre-processing steps. Extensive experiments demonstrate LensNet's robust performance and superior reconstruction quality compared to state-of-the-art methods, particularly in preserving high-frequency details and attenuating noise. The proposed framework establishes a novel convergence between physics-based modeling and data-driven learning, paving the way for more accurate, flexible, and practical lensless imaging solutions for applications ranging from miniature sensors to medical diagnostics. The link of code is https://github.com/baijiesong/Lensnet.

Figures

Figures reproduced from arXiv: 2505.01755 by the authors.

**Figure 2.** Visualized results on the DiffuserCam dataset. Each image shows the input image, the LensNet reconstruction result, and the [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** The figure illustrates the detailed structure of our model. The input image is measurement on the system and the result is clear [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (8 more)

**Figure 4.** Figure 4: The above shows the qualitative experimental results on MWDNs Dataset. Our method effectively achieves high-quality recon [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 5.** Figure 5: This figure presents the results of our ablation study, high [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Left image is input,and right image is target. The central [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]

**Figure 7.** Figure 7: Comprehensive visual comparison on MWDNs Dataset. (a) Input (b) ADMM , (c) APGD , (d) TikNet, (e) LenslessGAN,(f) UDN, [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

**Figure 8.** Figure 8: Comprehensive visual comparison on DiffuserCam Dataset. (a) ADMM , (b) APGD , (c) FlatNet, (d) UDN, (e) MWDN, (f) [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]

**Figure 9.** Figure 9: Comprehensive visual comparison on MWDNs Dataset. (a) Input (b) ADMM , (c) APGD , (d) TikNet, (e) LenslessGAN,(f) UDN, [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]

**Figure 10.** Figure 10: Comprehensive visual comparison on MWDNs Dataset. (a) Input (b) ADMM , (c) APGD , (d) TikNet, (e) LenslessGAN,(f) UDN, [PITH_FULL_IMAGE:figures/full_fig_p015_10.png]

**Figure 11.** Figure 11: Comprehensive visual comparison on DiffuserCam Dataset. (a) ADMM , (b) APGD , (c) FlatNet, (d) UDN, (e) MWDN, (f) [PITH_FULL_IMAGE:figures/full_fig_p016_11.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

Integrated Forward-Inverse Network for Lensless Image Reconstruction
cs.CV 2026-07 accept novelty 6.0 of 10

Interleaving differentiable forward projections with learnable inverse updates at every encoder–decoder scale yields state-of-the-art lensless image reconstruction and transfers to deblurring and holography.
CMAMRNet: A Contextual Mask-Aware Network Enhancing Mural Restoration Through Comprehensive Mask Guidance
cs.CV 2025-08 conditional novelty 5.0 of 10

A mask-aware transformer U-Net with mask-guided up/down-samplers and a co-feature aggregator reports state-of-the-art mural restoration on MuralDH and Dunhuang benchmarks.

Reference graph

Works this paper leans on

31 extracted references · 30 canonical work pages · cited by 2 Pith papers

[1]

Diffusercam: lensless single-exposure 3d imaging

[Antipa et al., 2017] Nick Antipa, Grace Kuo, Reinhard Heckel, Ben Mildenhall, Emrah Bostan, Ren Ng, and Laura Waller. Diffusercam: lensless single-exposure 3d imaging. Optica, 5(1):1–9,

work page 2017
[9]

Cross-view geo-localization via learning correspondence semantic similarity knowledge

[Chen et al., 2025] Guanli Chen, Guoheng Huang, Xiaochen Yuan, Xuhang Chen, Guo Zhong, and Chi-Man Pun. Cross-view geo-localization via learning correspondence semantic similarity knowledge. In MMM, pages 220–233,

work page 2025
[12]

Iman: An adaptive network for robust npc mortality prediction with missing modalities

[Huo et al., 2024] Yejing Huo, Guoheng Huang, Lianglun Cheng, Jianbin He, Xuhang Chen, Xiaochen Yuan, Guo Zhong, and Chi-Man Pun. Iman: An adaptive network for robust npc mortality prediction with missing modalities. In BIBM, pages 2074–2079,

work page 2024
[13]

Towards photorealistic reconstruction of highly multiplexed lensless images

[Khan et al., 2019] Salman S Khan, VR Adarsh, Vivek Boominathan, Jasper Tan, Ashok Veeraraghavan, and Kaushik Mitra. Towards photorealistic reconstruction of highly multiplexed lensless images. In ICCV, pages 7860– 7869,

work page 2019
[15]

Unrolled primal- dual networks for lensless cameras

[Kingshott et al., 2022] Oliver Kingshott, Nick Antipa, Em- rah Bostan, and Kaan Aks ¸it. Unrolled primal- dual networks for lensless cameras. Optics Express , 30(26):46324–46335,

work page 2022
[17]

Depth- aware test-time training for zero-shot video object seg- mentation

[Liu et al., 2024] Weihuang Liu, Xi Shen, Haolun Li, Xiuli Bi, Bo Liu, Chi-Man Pun, and Xiaodong Cun. Depth- aware test-time training for zero-shot video object seg- mentation. In CVPR, pages 19218–19227,

work page 2024
[18]

Devi- gnet: High-resolution vignetting removal via a dual aggre- gated fusion transformer with adaptive channel expansion

[Luo et al., 2024] Shenghong Luo, Xuhang Chen, Weiwen Chen, Zinuo Li, Shuqiang Wang, and Chi-Man Pun. Devi- gnet: High-resolution vignetting removal via a dual aggre- gated fusion transformer with adaptive channel expansion. In AAAI, pages 4000–4008,

work page 2024
[19]

Learned reconstructions for practical mask-based lensless imaging

[Monakhova et al., 2019] Kristina Monakhova, Joshua Yurt- sever, Grace Kuo, Nick Antipa, Kyrollos Yanny, and Laura Waller. Learned reconstructions for practical mask-based lensless imaging. Optics express, 27(20):28075–28090,

work page 2019

Show all 31 references

[20]

Introductory lectures on convex programming volume i: Basic course

[Nesterov, 1998] Yurii Nesterov. Introductory lectures on convex programming volume i: Basic course. Lecture notes, 3(4):5,

1998
[22]

Robust lensless image reconstruction via psf estimation

[Rego et al., 2021] Joshua D Rego, Karthik Kulkarni, and Suren Jayasuriya. Robust lensless image reconstruction via psf estimation. In WACV, pages 403–412,

2021
[26]

Extrapolation, interpola- tion, and smoothing of stationary time series: with engi- neering applications

[Wiener, 1949] Norbert Wiener. Extrapolation, interpola- tion, and smoothing of stationary time series: with engi- neering applications. The MIT press,

1949
[28]

Swift sampler: Efficient learning of sampler by 10 param- eters

[Yao et al., 2024] Jiawei Yao, Chuming Li, and Canran Xiao. Swift sampler: Efficient learning of sampler by 10 param- eters. arXiv preprint arXiv:2410.05578,

2024 arXiv
[29]

Robust reconstruction with deep learning to handle model mismatch in lensless imaging

[Zeng and Lam, 2021] Tianjiao Zeng and Edmund Y Lam. Robust reconstruction with deep learning to handle model mismatch in lensless imaging. TCI, 7:1080–1092,

2021
[30]

Smaformer: Synergistic multi- attention transformer for medical image segmentation

[Zheng et al., 2024] Fuchen Zheng, Xuhang Chen, Wei- huang Liu, Haolun Li, Yingtie Lei, Jiahui He, Chi-Man Pun, and Shounjun Zhou. Smaformer: Synergistic multi- attention transformer for medical image segmentation. In BIBM,

2024
[31]

Docdeshadower: Frequency-aware transformer for document shadow removal

[Zhou et al., 2024] Ziyang Zhou, Yingtie Lei, Xuhang Chen, Shenghong Luo, Wenjun Zhang, Chi-Man Pun, and Zhen Wang. Docdeshadower: Frequency-aware transformer for document shadow removal. In SMC, pages 2468–2473,

2024
[1949]

Tcia: A transformer-cnn model with illumination adaptation for enhancing cell im- age saliency and contrast

[Yang et al., 2025] Jietao Yang, Guoheng Huang, Yanzhang Luo, Xiaofeng Zhang, Xiaochen Yuan, Xuhang Chen, Chi- Man Pun, and Mu-yan Cai. Tcia: A transformer-cnn model with illumination adaptation for enhancing cell im- age saliency and contrast. TIM,

2025
[1998]

Harmonicnerf: Geometry- informed synthetic view augmentation for 3d scene re- construction in driving scenarios

[Pan et al., 2024] Xiaochao Pan, Jiawei Yao, Hongrui Kou, Tong Wu, and Canran Xiao. Harmonicnerf: Geometry- informed synthetic view augmentation for 3d scene re- construction in driving scenarios. In Proceedings of the 32nd ACM International Conference on Multimedia, pages 5987–5996,

2024
[2001]

An admm algorithm for a class of total variation regularized estimation problems

[Wahlberg et al., 2012] Bo Wahlberg, Stephen Boyd, Mari- ette Annergren, and Yang Wang. An admm algorithm for a class of total variation regularized estimation problems. IFAC Proceedings Volumes, 45(16):83–88,

2012
[2005]

Dual-hybrid at- tention network for specular highlight removal

[Guo et al., 2024] Xiaojiao Guo, Xuhang Chen, Shenghong Luo, Shuqiang Wang, and Chi-Man Pun. Dual-hybrid at- tention network for specular highlight removal. In ACM MM, pages 10173–10181,

2024
[2009]

Phlatcam: Designed phase-mask based thin lensless cam- era

[Boominathan et al., 2020] Vivek Boominathan, Jesse K Adams, Jacob T Robinson, and Ashok Veeraraghavan. Phlatcam: Designed phase-mask based thin lensless cam- era. TPAMI, 42(7):1618–1629,

2020
[2011]

Phocolens: Pho- torealistic and consistent reconstruction in lensless imag- ing

[Cai et al., 2024] Xin Cai, Zhiyuan You, Hailong Zhang, Wentao Liu, Jinwei Gu, and Tianfan Xue. Phocolens: Pho- torealistic and consistent reconstruction in lensless imag- ing. arXiv,

2024
[2012]

Global convergence of admm in nonconvex nonsmooth optimization

[Wang et al., 2019] Yu Wang, Wotao Yin, and Jinshan Zeng. Global convergence of admm in nonconvex nonsmooth optimization. Journal of Scientific Computing, 78:29–63,

2019
[2015]

Flatcam: Thin, lensless cameras using coded aperture and computation

[Asif et al., 2016] M Salman Asif, Ali Ayremlou, Aswin Sankaranarayanan, Ashok Veeraraghavan, and Richard G Baraniuk. Flatcam: Thin, lensless cameras using coded aperture and computation. TCI, 3(3):384–397,

2016
[2017]

Flatcam: Replacing lenses with masks and computation

[Asif et al., 2015] M Salman Asif, Ali Ayremlou, Ashok Veeraraghavan, Richard Baraniuk, and Aswin Sankara- narayanan. Flatcam: Replacing lenses with masks and computation. In ICCV Workshop, pages 663–666. IEEE,

2015
[2019]

Flatnet: Towards photorealistic scene reconstruc- tion from lensless measurements

[Khan et al., 2020] Salman Siddique Khan, Varun Sundar, Vivek Boominathan, Ashok Veeraraghavan, and Kaushik Mitra. Flatnet: Towards photorealistic scene reconstruc- tion from lensless measurements. TPAMI, 44(4):1934– 1948,

2020
[2020]

Distributed op- timization and statistical learning via the alternating direc- tion method of multipliers

[Boyd et al., 2011] Stephen Boyd, Neal Parikh, Eric Chu, Borja Peleato, Jonathan Eckstein, et al. Distributed op- timization and statistical learning via the alternating direc- tion method of multipliers. Foundations and Trends® in Machine learning, 3(1):1–122,

2011
[2021]

Thin observation module by bound op- tics (tombo): concept and experimental verification

[Tanida et al., 2001] Jun Tanida, Tomoya Kumagai, Kenji Yamada, Shigehiro Miyatake, Kouichi Ishida, Takashi Morimoto, Noriyuki Kondou, Daisuke Miyazaki, and Yoshiki Ichioka. Thin observation module by bound op- tics (tombo): concept and experimental verification. Ap- plied opt...

2001
[2022]

Accelerated proximal gradient methods for nonconvex programming

[Li and Lin, 2015] Huan Li and Zhouchen Lin. Accelerated proximal gradient methods for nonconvex programming. NeurIPS, 28,

2015
[2023]

Medprompt: Cross-modal prompting for multi-task medical image translation

[Chen et al., 2024] Xuhang Chen, Chi-Man Pun, and Shuqiang Wang. Medprompt: Cross-modal prompting for multi-task medical image translation. In PRCV, pages 61–75,

2024
[2024]

Brain diffuser: An end-to-end brain image to brain network pipeline

[Chen et al., 2023] Xuhang Chen, Baiying Lei, Chi-Man Pun, and Shuqiang Wang. Brain diffuser: An end-to-end brain image to brain network pipeline. In PRCV, pages 16–26,

2023
[2025]

Introduction to Fourier optics

[Goodman, 2005] Joseph W Goodman. Introduction to Fourier optics. Roberts and Company publishers,

2005

Pith tools

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

[1] [1]

Diffusercam: lensless single-exposure 3d imaging

[Antipa et al., 2017] Nick Antipa, Grace Kuo, Reinhard Heckel, Ben Mildenhall, Emrah Bostan, Ren Ng, and Laura Waller. Diffusercam: lensless single-exposure 3d imaging. Optica, 5(1):1–9,

work page 2017

[2] [9]

Cross-view geo-localization via learning correspondence semantic similarity knowledge

[Chen et al., 2025] Guanli Chen, Guoheng Huang, Xiaochen Yuan, Xuhang Chen, Guo Zhong, and Chi-Man Pun. Cross-view geo-localization via learning correspondence semantic similarity knowledge. In MMM, pages 220–233,

work page 2025

[3] [12]

Iman: An adaptive network for robust npc mortality prediction with missing modalities

[Huo et al., 2024] Yejing Huo, Guoheng Huang, Lianglun Cheng, Jianbin He, Xuhang Chen, Xiaochen Yuan, Guo Zhong, and Chi-Man Pun. Iman: An adaptive network for robust npc mortality prediction with missing modalities. In BIBM, pages 2074–2079,

work page 2024

[4] [13]

Towards photorealistic reconstruction of highly multiplexed lensless images

[Khan et al., 2019] Salman S Khan, VR Adarsh, Vivek Boominathan, Jasper Tan, Ashok Veeraraghavan, and Kaushik Mitra. Towards photorealistic reconstruction of highly multiplexed lensless images. In ICCV, pages 7860– 7869,

work page 2019

[5] [15]

Unrolled primal- dual networks for lensless cameras

[Kingshott et al., 2022] Oliver Kingshott, Nick Antipa, Em- rah Bostan, and Kaan Aks ¸it. Unrolled primal- dual networks for lensless cameras. Optics Express , 30(26):46324–46335,

work page 2022

[6] [17]

Depth- aware test-time training for zero-shot video object seg- mentation

[Liu et al., 2024] Weihuang Liu, Xi Shen, Haolun Li, Xiuli Bi, Bo Liu, Chi-Man Pun, and Xiaodong Cun. Depth- aware test-time training for zero-shot video object seg- mentation. In CVPR, pages 19218–19227,

work page 2024

[7] [18]

Devi- gnet: High-resolution vignetting removal via a dual aggre- gated fusion transformer with adaptive channel expansion

[Luo et al., 2024] Shenghong Luo, Xuhang Chen, Weiwen Chen, Zinuo Li, Shuqiang Wang, and Chi-Man Pun. Devi- gnet: High-resolution vignetting removal via a dual aggre- gated fusion transformer with adaptive channel expansion. In AAAI, pages 4000–4008,

work page 2024

[8] [19]

Learned reconstructions for practical mask-based lensless imaging

[Monakhova et al., 2019] Kristina Monakhova, Joshua Yurt- sever, Grace Kuo, Nick Antipa, Kyrollos Yanny, and Laura Waller. Learned reconstructions for practical mask-based lensless imaging. Optics express, 27(20):28075–28090,

work page 2019