Pith. sign in

REVIEW 3 major objections 4 minor 68 references

This paper argues that the 1D-to-2D 'lift' in image-free single-pixel sensing is the decisive design axis, and that a content-adaptive lift beats even task-adapted reconstruction once measurement noise is present.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 05:51 UTC pith:ON6CWIH6

load-bearing objection A genuinely useful reframing of the lift in image-free single-pixel sensing, with a solid architecture study and an honest mechanism analysis; the central noise-reversal claim needs one missing control before it should be taken at face value. the 3 major comments →

arxiv 2607.22077 v1 pith:ON6CWIH6 submitted 2026-07-24 eess.IV cs.CVphysics.optics

The Lift Spectrum: How Measurement-to-Space Adaptivity Shapes Robustness in Image-Free Single-Pixel Sensing

Yuyuan Han , Jingwei Li , Long Qiu , Chong Wang , Wenxuan Hao , Jiangyu Han , Xinyu Yao , Yuchen He
show 3 more authors
Hui Chen Jianbin Liu Huaibin Zheng
This is my paper
classification eess.IV cs.CVphysics.optics
keywords single-pixel sensingimage-free inferencelift spectrumcross-attention lifttask-prioritized loss schedulingmeasurement noise robustnessHadamard patternsimage segmentation
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Single-pixel sensing compresses a scene into a short sequence of coded measurements; the paper's claim is that the hard part of image-free segmentation is neither the measurements nor the decoder, but the 'lift' — the map from a 1D sequence to a 2D feature map. It orders methods by how adaptive this lift is: a fixed physics inverse, a learned static projection, or a content-adaptive retrieval, and argues that position on this spectrum predicts which method wins as sampling drops and noise rises. The proposed STSF+TPLS, built on a GRU encoder, a cross-attention lift, and a scheduled reconstruction prior, improves foreground mean intersection-over-union by +3.2 to +9.9 percentage points over the previous image-free baseline at 3.13% sampling and holds accuracy to 0.39%. In the noiseless limit the reconstruct-then-segment route is more accurate; under calibrated 20 dB measurement noise image-free inference overtakes it, because the reconstruction pipeline amplifies the identical noise 20–70× before its segmenter reads it. Each lift region also fails characteristically — collapse, imprinting, coarsening — so choosing a lift is choosing a failure mode.

Core claim

The central claim is that under extreme single-pixel compression, reconstructing the scene is the wrong inference target but the right learning signal. Removing reconstruction relocates the difficulty to the lift, the step that maps a 1D measurement sequence to a 2D feature map. Ordering lifts by adaptivity — fixed-physics inverse, learned static projection, content-adaptive retrieval — predicts behavior as the acquisition degrades. The proposed STSF+TPLS (GRU encoder, cross-attention lift, U-Net++ decoder, scheduled reconstruction prior) improves foreground mIoU by +3.2 to +9.9 percentage points over the prior image-free baseline at 3.13% sampling and plateaus to 0.39%. In the noiseless lim

What carries the argument

The central object is the lift, the map from the 1D measurement sequence to a 2D representation. STSF instantiates the content-adaptive end: 512 measurements split into 64 tokens of length 8, encoded by a three-layer bidirectional GRU, then read by a 16×16 grid of learnable spatial queries through four cross-attention blocks, with a ConvNeXt and U-Net++ decoder emitting the mask. Task-prioritized loss scheduling (TPLS) anneals an auxiliary reconstruction loss (L1+MSE+DSSIM) from weight 0.9 to 0.1 so the physical prior leads early and the segmentation loss takes over. Eq. 7 — δ_TA/δ_IF = (√M/N)(R_IF/R_TA) — carries the noise-reversal argument: the measurement-domain range is anchored by the l

Load-bearing premise

The operating-regime map is measured under one fixed pattern ordering and per-sample normalization, and the temporal encoder is chosen on the belief that reconstruction quality ranks segmentation-relevant structure; if those change, the reported ordering of methods could shift.

What would settle it

Re-run the 20 dB noise sweep after replacing per-sample min–max normalization with a single global normalization, or after switching to sequency-ordered Hadamard patterns; if image-free inference still overtakes reconstruction, the DC-anchored-range mechanism is not the cause.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • A noiseless benchmark systematically favors reconstruct-then-segment; robustness claims must state the operating SNR.
  • At 20 dB and 3.13% sampling, image-free STSF+TPLS beats the strongest clean-trained reconstruction baseline by +4.7, +21.7, and +3.1 pp foreground mIoU on Carvana, MNIST, and WBC.
  • Accuracy stays object-shaped down to 0.39% sampling, within 5.7, 2.0, and 3.4 pp of the 3.13% result.
  • Each lift region has a characteristic failure — collapse for the fixed-physics inverse, template imprinting for the learned-static projection, coarsening for the content-adaptive lift — so the design choice is also a choice of failure mode.
  • On real single-pixel hardware, image-free transfer without fine-tuning reproduces the simulated reversal on the reliable-ground-truth target and is stable across repeated acquisitions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • (Beyond the paper) If the noise-amplification mechanism is general, any task whose input is a min–max-normalized reconstruction inherits the same 20–70× relative perturbation; detection, classification, and pose estimation are testable next targets.
  • (Beyond the paper) The reversal may depend on the deliberately unoptimized pattern ordering; substituting sequency-ordered Hadamard or learned illumination would show whether the principle is about normalization placement or about pattern choice.
  • (Beyond the paper) TPLS can be read as a curriculum in which reconstruction is scaffolding; the same scheduled-prior trick could transfer to other ill-posed inverse tasks with a cheap auxiliary objective.
  • (Beyond the paper) The three failure signatures give a field diagnostic: masks that vanish under noise signal a fixed-physics lift, masks that stop changing signal a static lift, and masks that dilate signal a content-adaptive lift.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes a 'lift spectrum' for image-free single-pixel segmentation, ordering methods by how adaptively they map a 1D measurement sequence to the 2D representation read by a segmentation decoder: a fixed-physics inverse (reconstruct-then-segment), a learned static projection (SPIFS-style), and the proposed content-adaptive cross-attention retrieval. The proposed STSF network uses a GRU temporal encoder, a cross-attention 1D-to-2D lift, and a U-Net++ decoder, trained with task-prioritized loss scheduling (TPLS), an auxiliary reconstruction prior whose weight is annealed. The paper reports consistent gains over a re-implemented SPIFS baseline at 3.13% sampling, a flat accuracy plateau down to 0.39% sampling, and—under calibrated measurement noise—a reversal in which image-free inference beats a task-adapted (TA) reconstruct-then-segment baseline. The reversal is attributed to a measured 20–70× amplification of the identical measurement noise before it reaches the TA segmenter, formalized in Eq. (7). The paper also characterizes failure modes and demonstrates a single-pixel hardware proof of concept.

Significance. If the central robustness claim holds, the paper is a useful contribution: it identifies the normalization point and the noise-amplification mechanism as a design determinant, provides a principled map of when to use reconstruction versus image-free inference, and ships code and pretrained weights. The paper is unusually honest about its own limitations: single-seed studies are disclosed as such, the SPIFS baseline is disclosed as a re-implementation without learned illumination, the parameter gap versus SPIFS is stated, and the noise-reversal mechanism is tested with a noise-augmented retraining intervention that supports the distribution-shift account. The parameter-matched lift ablation and the multi-seed noise sweep are carefully executed. However, the central robustness claim is tied to a specific, untested normalization choice in the reconstruction arm, and the headline comparison to the 'prior image-free baseline' rests on a weakened re-implementation. These issues are load-bearing for the abstract and conclusions and require additional experiments or carefully qualified claims.

major comments (3)
  1. [§IV-D, Eq. (7), Supplementary S3] The measured 20–70× noise amplification and the resulting reversal depend on the TA baseline's per-image min–max normalization being 'DC-blind.' Eq. (7) shows that the range factor R_IF/R_TA carries the entire amplification, and the text explains that this is because per-image min–max discards the spatially uniform DC plane in the reconstruction. The paper never varies the reconstruction arm's normalization. A physically motivated alternative—dividing the reconstruction by the measured DC coefficient (s0/N), or using a fixed/global normalization fit on clean reconstructions—would change R_TA and could substantially reduce the amplification (to roughly sqrt(M) in the idealized first-order calculation). If TA accuracy at 20 dB recovers under such a control, the claim that image-free inference is intrinsically more robust, as stated in the abstract and conclusion, is an artifact of the base
  2. [§IV-A, Fig. 3, Supplementary S1] The headline gain against the 'prior image-free baseline' SPIFS is measured against a re-implementation that, as disclosed, omits the original learned illumination, drops the original two-stage pretraining, changes the segmentation loss from MSE to Dice+CE, and uses a fixed Hadamard acquisition that may understate SPIFS. The paper also notes that STSF has ~22% more parameters than the SPIFS re-implementation. These are acknowledged in the text, but the abstract and conclusion state the +3.2 to +9.9 pp improvement as a result against 'the prior image-free baseline' without these caveats. The comparison is load-bearing for the first half of the central claim. Please either run the original SPIFS (or a faithful reproduction with learned illumination and its pretraining) under the same acquisition, or explicitly reframe the claim as improvement over a fixed-acquisition, parameter-unmatched r
  3. [§IV-B, Table I] The temporal encoder is selected using the single-run reconstruction probe of Table I, with GRU chosen over LSTM by very small margins on CelebA (PSNR 15.69 vs 15.67; SSIM 0.453 vs 0.454, where LSTM leads) and by ~0.1 dB on Fashion-MNIST. The table itself states that every cell is a single training run and that absolute values carry no seed statistics. A single-run ranking with margins this thin is not a robust basis for a central architectural choice, even if the protocol is identical across architectures. Please add at least three seeds per architecture for the key comparison (GRU vs LSTM vs FC) at 3.13% sampling, or provide an end-to-end segmentation ablation of the encoder choice, to show that the GRU selection is not an artifact of seed noise.
minor comments (4)
  1. [Abstract and §IV-D] The abstract says image-free inference 'overtakes' the reconstruction baseline under measurement noise 'for a measured reason,' with no mention of the normalization sensitivity identified in Eq. (7). Consider adding a qualifier such as 'under the tested per-image min–max reconstruction normalization' to avoid overclaiming until the control is run.
  2. [§IV-D, Fig. 6(a–c)] The 32× sampling-rate plateau is based on a single seed at most rates; the paper discloses this and re-runs only the narrowest margin. This is acceptable if clearly labeled in the figure and abstract, but the abstract's unqualified 'plateaus down to 0.39%' should be softened to 'in single-seed sweeps, re-confirmed at the narrowest margin.'
  3. [§IV-F, Figs. 8–9] The hardware 'reversal' is a proof of concept on one target with a known ground truth. The paper already states this, but the conclusion's phrasing 'the real bench echoes the same signature' may be stronger than warranted; suggest 'is consistent with' rather than 'echoes.'
  4. [Supplementary S1] Minor typographical / notation issues: the loss weight λ2 in Eq. (4) is used for the MSE term, which is fine, but in Supplementary S1 the composite reconstruction loss for the probe is described with weights 0.5/0.3/0.2 for L1/MSE/SSIM, while main-text Eq. (4) uses 0.3/0.3/0.4. Clarify that the probe and the segmentation auxiliary loss use different weights.

Circularity Check

0 steps flagged

No significant circularity: the central derivation and claims rest on external benchmarks and end-to-end measurements; only peripheral self-citations appear.

full rationale

The paper's central comparisons are measured against external baselines (SPIFS, TA-HSI/CS) under a shared fixed acquisition, not fitted constants renamed as predictions. The noise-reversal argument (Sec. IV-D, Eq. 7) is a mechanistic decomposition of a measured effect: the range ratio R_IF/R_TA is measured directly, the 20–70× amplification is measured end-to-end, and the first-order formula is checked against that measurement rather than used to define it. The TA baseline's per-image min–max normalization is disclosed; the absence of a DC-anchored reconstruction-normalization control is a legitimate robustness concern, but it does not make the derivation circular, because the claim concerns the pipeline as instantiated. The temporal-encoder selection is explicitly labeled a 'stated proxy' (Sec. IV-B) and is backed by end-to-end ablations (Table II, Fig. 4). Self-citations ([6], [21], [26]) appear only in background context and are not load-bearing; no uniqueness theorem is imported, and no ansatz is smuggled in via self-citation. Limitations (single-seed sweeps, fixed acquisition, proof-of-concept hardware) are disclosed in the manuscript, further reducing any concern that a predicted result is equivalent to its inputs by construction.

Axiom & Free-Parameter Ledger

3 free parameters · 5 axioms · 0 invented entities

The paper's central claims rest on empirical comparisons under one fixed Hadamard acquisition. The free parameters are training-schedule and loss hyperparameters; the main axioms are the linear forward model, orthogonal Hadamard rows, the encoder-selection proxy, and the noise-calibration/normalization assumptions underlying the reversal mechanism. No new physical entities or forces are introduced.

free parameters (3)
  • TPLS phase fractions (r1, r2) = Carvana/MNIST (0.3, 0.3); WBC (0.6, 0.3)
    Per-dataset phase boundaries chosen by validation; the TPLS benefit depends on these boundaries.
  • TPLS weight levels (α, β) = Phase 1: 0.9/0.1; Phase 2: 0.5/0.5; Phase 3: 0.1/0.9
    Hand-chosen schedule levels; the fixed-AL ablation uses the same middle weighting, so the schedule values are load-bearing for the TPLS comparison.
  • Auxiliary and segmentation loss weights = λ1=λ2=0.3, λssim=0.4; BCE/Dice = 0.5/0.5
    Hand-chosen composition of the auxiliary reconstruction loss and the segmentation loss; no sensitivity analysis is reported.
axioms (5)
  • domain assumption Linear acquisition model s = Φx + n (Eq. 1)
    The entire simulation and hardware interpretation assume single-pixel acquisition is a linear projection of the scene plus measurement noise.
  • standard math Retained Sylvester-Hadamard rows satisfy ΦΦᵀ = N·I (Supp. Eq. S2)
    Used to argue the reconstruction adjoint is well-conditioned and does not amplify noise per mode.
  • ad hoc to paper Reconstruction fidelity is a valid proxy for segmentation-relevant encoder quality (Sec. IV-B)
    The temporal encoder (GRU) is selected on a reconstruction benchmark; this transfers to segmentation only through a stated proxy, not a proof.
  • domain assumption AC-coupled Gaussian noise calibration σ = std(s_AC)·10^(−SNR/20) represents deployment measurement noise (Sec. IV-D)
    The noise-reversal claim is established under this calibration in simulation; the supplementary Poisson sweep is a partial robustness check, not a full detector-noise model.
  • standard math Per-image min–max normalization is shift-invariant, so the DC plane does not anchor the reconstruction range (Sec. IV-D, Supp. S3)
    This is the key mechanism behind the measured 20–70× noise amplification; the claim depends on this normalization behavior being preserved in deployment.

pith-pipeline@v1.3.0-alltime-deepseek · 35120 in / 11759 out tokens · 130936 ms · 2026-08-01T05:51:01.266371+00:00 · methodology

0 comments
read the original abstract

Single-pixel sensing encodes a scene as a short sequence of coded measurements, and image-free methods infer the task directly from that sequence. Removing reconstruction does not remove the difficulty: it relocates it to the lift, the map from 1D measurements to a 2D representation, which prior work treats as a trivial reshape. We recast the lift as the central design axis of image-free sensing and order methods by how strongly it adapts to its input: a fixed-physics inverse (reconstruct-then-segment), a learned static projection, or a content-adaptive retrieval; position on this lift spectrum predicts behavior as acquisition degrades. The spatiotemporal soft-fusion (STSF) network pairs a probe-selected recurrent encoder with a cross-attention lift chosen by a parameter-matched ablation, ahead of its U-Net++ decoder, and trains under task-prioritized loss scheduling (TPLS), a scheduled reconstruction prior. In simulation, STSF+TPLS surpasses the prior image-free baseline on three datasets at 3.13% sampling (+3.2 to +9.9 pp foreground mIoU) and plateaus down to 0.39%. The strongest clean-trained reconstruct-then-segment baseline wins the noiseless limit, but under calibrated measurement noise image-free inference overtakes it, for a measured reason: the reconstruction pipeline amplifies the identical measurement noise before its segmenter reads it. Each region fails in its own signature: collapse, imprinting, or coarsening. STSF+TPLS transfers without fine-tuning to a real single-pixel bench as a proof of concept, at about 14 ms per mask. Charting the lift turns a scattered design space into a map of which lift to deploy at each operating point. Code and pretrained weights: https://github.com/Hanyuyuan6/STSF-TPLS

Figures

Figures reproduced from arXiv: 2607.22077 by Chong Wang, Huaibin Zheng, Hui Chen, Jianbin Liu, Jiangyu Han, Jingwei Li, Long Qiu, Wenxuan Hao, Xinyu Yao, Yuchen He, Yuyuan Han.

Figure 1
Figure 1. Figure 1: Proposed STSF+TPLS architecture for image-free segmentation from a 1D single-pixel measurement sequence. (a) Acquisition and temporal encoding: [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Validation foreground mDice on Carvana under the three-phase schedule of Eq. (6) (phase boundaries dashed, phases shaded; the as-run loss weights follow the schedule exactly; the schedule itself is diagrammed in [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Foreground segmentation accuracy of the proposed STSF+TPLS [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Cumulative component ablation on the Carvana test set (foreground [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The reversal, qualitatively: segmentations at a 20 dB measurement SNR ( [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Operating regime across sampling rate and measurement noise; foreground mIoU at [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Quantitative failure signatures across the lift spectrum (MNIST, [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Real single-pixel bench, two valid physical targets at 3.13% sampling. (a, b) Acquisition: a projector displays a held-out test image on a matte screen; [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Sim-to-real on the physical bench: MNIST digit-1, the reliable-ground [PITH_FULL_IMAGE:figures/full_fig_p011_9.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

68 extracted references · 2 linked inside Pith

  1. [1]

    T. B. Pittman, Y. H. Shih, D. V. Strekalov, and A. V. Sergienko, ``Optical imaging by means of two-photon quantum entanglement,'' Phys. Rev. A, vol. 52, no. 5, p. R3429, 1995

    1995

  2. [2]

    M. F. Duarte et al., ``Single-pixel imaging via compressive sampling,'' IEEE Signal Process. Mag., vol. 25, no. 2, pp. 83--91, 2008

    2008

  3. [3]

    Rousset, N

    F. Rousset, N. Ducros, A. Farina, G. Valentini, C. d'Andrea, and F. Peyrin, ``Adaptive basis scan by wavelet prediction for single-pixel imaging,'' IEEE Trans. Comput. Imag., vol. 3, no. 1, pp. 36--46, 2017

    2017

  4. [4]

    Sun et al., `` 3D computational imaging with single-pixel detectors,'' Science, vol

    B. Sun et al., `` 3D computational imaging with single-pixel detectors,'' Science, vol. 340, no. 6134, pp. 844--847, 2013

    2013

  5. [5]

    Xu, W.-T

    Y.-K. Xu, W.-T. Liu, E.-F. Zhang, Q. Li, H.-Y. Dai, and P.-X. Chen, ``Is ghost imaging intrinsically more powerful against scattering?'' Opt. Express, vol. 23, no. 26, pp. 32\,993--33\,000, 2015

    2015

  6. [6]

    Han et al., ``Generalized sparse H adamard single-pixel imaging,'' Opt

    Y. Han et al., ``Generalized sparse H adamard single-pixel imaging,'' Opt. Express, vol. 33, no. 20, pp. 43\,324--43\,341, 2025

    2025

  7. [7]

    Wang and S

    L. Wang and S. Zhao, ``Fast reconstructed and high-quality ghost imaging with fast W alsh-- H adamard transform,'' Photon. Res., vol. 4, no. 6, pp. 240--244, 2016

    2016

  8. [8]

    Zhang, X

    Z. Zhang, X. Ma, and J. Zhong, ``Single-pixel imaging by means of F ourier spectrum acquisition,'' Nat. Commun., vol. 6, no. 1, p. 6225, 2015

    2015

  9. [9]

    R. I. Khakimov et al., ``Ghost imaging with atoms,'' Nature, vol. 540, no. 7631, pp. 100--103, 2016

    2016

  10. [10]

    Li et al., ``Dual-color terahertz spatial light modulator for single-pixel imaging,'' Light Sci

    W. Li et al., ``Dual-color terahertz spatial light modulator for single-pixel imaging,'' Light Sci. Appl., vol. 11, no. 1, p. 191, 2022

    2022

  11. [11]

    L. Peng, S. Xie, H. Lu, and L. Bian, ``Large-scale single-pixel imaging and sensing,'' Adv. Photon. Nexus, vol. 4, no. 2, p. 026010, 2025

    2025

  12. [12]

    L. Peng, S. Xie, T. Qin, L. Cao, and L. Bian, ``Image-free single-pixel object detection,'' Opt. Lett., vol. 48, no. 10, pp. 2527--2530, 2023

    2023

  13. [13]

    Song et al., ``Advances and challenges of single-pixel imaging based on deep learning,'' Laser Photon

    K. Song et al., ``Advances and challenges of single-pixel imaging based on deep learning,'' Laser Photon. Rev., vol. 19, no. 7, p. 2401397, 2025

    2025

  14. [14]

    Kulkarni and P

    K. Kulkarni and P. Turaga, ``Reconstruction-free action inference from compressive imagers,'' IEEE Trans. Pattern Anal. Mach. Intell., vol. 38, no. 4, pp. 772--784, 2016

    2016

  15. [15]

    Lohit, K

    S. Lohit, K. Kulkarni, and P. Turaga, ``Direct inference on compressive measurements using convolutional neural networks,'' in IEEE International Conference on Image Processing (ICIP), 2016, pp. 1913--1917

    2016

  16. [16]

    H. Fu, L. Bian, and J. Zhang, ``Single-pixel sensing with optimal binarized modulation,'' Opt. Lett., vol. 45, no. 11, pp. 3111--3114, 2020

    2020

  17. [17]

    H. Liu, L. Bian, and J. Zhang, ``Image-free single-pixel segmentation,'' Opt. Laser Technol., vol. 157, p. 108600, 2023

    2023

  18. [18]

    Z. Zhou, M. M. Rahman Siddiquee, N. Tajbakhsh, and J. Liang, `` UNet ++: A nested U-Net architecture for medical image segmentation,'' in International Workshop on Deep Learning in Medical Image Analysis, 2018, pp. 3--11

    2018

  19. [19]

    G. E. Karniadakis, I. G. Kevrekidis, L. Lu, P. Perdikaris, S. Wang, and L. Yang, ``Physics-informed machine learning,'' Nat. Rev. Phys., vol. 3, no. 6, pp. 422--440, 2021

    2021

  20. [20]

    Lyu et al., ``Deep-learning-based ghost imaging,'' Sci

    M. Lyu et al., ``Deep-learning-based ghost imaging,'' Sci. Rep., vol. 7, no. 1, p. 17865, 2017

    2017

  21. [21]

    He et al., ``Ghost imaging based on deep learning,'' Sci

    Y. He et al., ``Ghost imaging based on deep learning,'' Sci. Rep., vol. 8, no. 1, p. 6469, 2018

    2018

  22. [22]

    C. F. Higham, R. Murray-Smith, M. J. Padgett, and M. P. Edgar, ``Deep learning for real-time single-pixel video,'' Sci. Rep., vol. 8, no. 1, p. 2369, 2018

    2018

  23. [23]

    F. Wang, H. Wang, H. Wang, G. Li, and G. Situ, ``Learning from simulation: An end-to-end deep-learning approach for computational ghost imaging,'' Opt. Express, vol. 27, no. 18, pp. 25\,560--25\,572, 2019

    2019

  24. [24]

    Hoshi, T

    I. Hoshi, T. Shimobaba, T. Kakue, and T. Ito, ``Single-pixel imaging using a recurrent neural network combined with convolutional layers,'' Opt. Express, vol. 28, no. 23, pp. 34\,069--34\,078, 2020

    2020

  25. [25]

    F. Wang, C. Wang, C. Deng, S. Han, and G. Situ, ``Single-pixel imaging using physics enhanced deep learning,'' Photon. Res., vol. 10, no. 1, pp. 104--110, 2022

    2022

  26. [26]

    He et al., `` ADMMNet -based deep unrolling method for ghost imaging,'' IEEE Trans

    Y. He et al., `` ADMMNet -based deep unrolling method for ghost imaging,'' IEEE Trans. Comput. Imag., vol. 10, pp. 233--245, 2024

    2024

  27. [27]

    Zhang, C

    X. Zhang, C. Deng, C. Wang, F. Wang, and G. Situ, `` VGenNet : variable generative prior enhanced single pixel imaging,'' ACS Photonics, vol. 10, no. 7, pp. 2363--2373, 2023

    2023

  28. [28]

    G. Qu, P. Wang, and X. Yuan, ``Dual-scale transformer for large-scale single-pixel imaging,'' in IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 25\,327--25\,337

    2024

  29. [29]

    Sholokhov et al., ``Single-pixel imaging of spatio-temporal flows using differentiable latent dynamics,'' IEEE Trans

    A. Sholokhov et al., ``Single-pixel imaging of spatio-temporal flows using differentiable latent dynamics,'' IEEE Trans. Comput. Imag., vol. 10, pp. 1124--1138, 2024

    2024

  30. [30]

    P. Wang, L. Wang, G. Qu, X. Wang, Y. Zhang, and X. Yuan, ``Proximal algorithm unrolling: Flexible and efficient reconstruction networks for single-pixel imaging,'' in IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025

    2025

  31. [31]

    Ota et al., ``Ghost cytometry,'' Science, vol

    S. Ota et al., ``Ghost cytometry,'' Science, vol. 360, no. 6394, pp. 1246--1251, 2018

    2018

  32. [32]

    Li et al., ``Spectrally encoded single-pixel machine vision using diffractive networks,'' Sci

    J. Li et al., ``Spectrally encoded single-pixel machine vision using diffractive networks,'' Sci. Adv., vol. 7, no. 13, p. eabd7690, 2021

    2021

  33. [33]

    Yang et al., `` SP-ILC : Concurrent single-pixel imaging, object location, and classification by deep learning,'' Photonics, vol

    Z. Yang et al., `` SP-ILC : Concurrent single-pixel imaging, object location, and classification by deep learning,'' Photonics, vol. 8, no. 9, p. 400, 2021

    2021

  34. [34]

    Zhang, J

    Z. Zhang, J. Ye, Q. Deng, and J. Zhong, ``Image-free real-time detection and tracking of fast moving object using a single-pixel detector,'' Opt. Express, vol. 27, no. 24, pp. 35\,394--35\,401, 2019

    2019

  35. [35]

    H. Yue, J. Guo, X. Yin, Y. Zhang, B. Wen, and C. Li, ``Salient object detection toward single-pixel imaging,'' IEEE Trans. Circuits Syst. Video Technol., vol. 34, no. 1, pp. 235--247, 2024

    2024

  36. [36]

    Meng et al., ``Image-free Hu invariant moment measurement by single-pixel detection,'' Opt

    W. Meng et al., ``Image-free Hu invariant moment measurement by single-pixel detection,'' Opt. Laser Technol., vol. 181, p. 111581, 2025

    2025

  37. [37]

    Liu et al., ``Image-free single-pixel semantic segmentation for complex scene based on multi-scale U-Net ,'' Chin

    T. Liu et al., ``Image-free single-pixel semantic segmentation for complex scene based on multi-scale U-Net ,'' Chin. Phys. B, vol. 35, no. 1, p. 014202, 2026

    2026

  38. [38]

    Y. Wang, K. Hu, P. Dai, and J. Liu, ``Image-free single-pixel sensing for human pose estimation and parameter-efficient fine-tuning,'' Opt. Lett., vol. 51, no. 7, pp. 1875--1878, 2026

    2026

  39. [39]

    X. Zhan, H. Lu, R. Yan, and L. Bian, ``Global-optimal semi-supervised learning for single-pixel image-free sensing,'' Opt. Lett., vol. 49, no. 3, pp. 682--685, 2024

    2024

  40. [40]

    Ji, Y.-C

    P.-C. Ji, Y.-C. Wang, H.-J. Zhang, Z.-H. Yang, and Y.-J. Yu, ``Image-free tracking of single-pixel detection in dynamic complex scenes,'' Opt. Express, vol. 33, no. 21, p. 44497, 2025

    2025

  41. [41]

    Barbastathis, A

    G. Barbastathis, A. Ozcan, and G. Situ, ``On the use of deep learning for computational imaging,'' Optica, vol. 6, no. 8, pp. 921--943, 2019

    2019

  42. [42]

    Qin et al., ``Joint motion estimation and segmentation from undersampled cardiac MR image,'' in Machine Learning for Medical Image Reconstruction (MLMIR), MICCAI Workshop, 2018, pp

    C. Qin et al., ``Joint motion estimation and segmentation from undersampled cardiac MR image,'' in Machine Learning for Medical Image Reconstruction (MLMIR), MICCAI Workshop, 2018, pp. 55--63

    2018

  43. [43]

    L. Sun, Z. Fan, X. Ding, Y. Huang, and J. Paisley, ``Joint CS-MRI reconstruction and segmentation with a unified deep network,'' in Information Processing in Medical Imaging (IPMI), 2019, pp. 492--504

    2019

  44. [44]

    Antun, F

    V. Antun, F. Renna, C. Poon, B. Adcock, and A. C. Hansen, ``On instabilities of deep learning in image reconstruction and the potential costs of AI ,'' Proc. Natl. Acad. Sci. USA, vol. 117, no. 48, pp. 30\,088--30\,095, 2020

    2020

  45. [45]

    Bhadra, V

    S. Bhadra, V. A. Kelkar, F. J. Brooks, and M. A. Anastasio, ``On hallucinations in tomographic image reconstruction,'' IEEE Trans. Med. Imag., vol. 40, no. 11, pp. 3249--3260, 2021

    2021

  46. [46]

    Kabkab, P

    M. Kabkab, P. Samangouei, and R. Chellappa, ``Task-aware compressed sensing with generative adversarial networks,'' in AAAI Conference on Artificial Intelligence (AAAI), vol. 32, no. 1, 2018

    2018

  47. [47]

    D. Liu, B. Wen, J. Jiao, X. Liu, Z. Wang, and T. S. Huang, ``Connecting image denoising and high-level vision tasks via deep learning,'' IEEE Trans. Image Process., vol. 29, pp. 3695--3706, 2020

    2020

  48. [48]

    Mou and J

    C. Mou and J. Zhang, `` TransCL : Transformer makes strong and flexible compressive learning,'' IEEE Trans. Pattern Anal. Mach. Intell., vol. 45, no. 4, pp. 5236--5251, 2023

    2023

  49. [49]

    E. L. Allgower and K. Georg, Introduction to Numerical Continuation Methods. 1em plus 0.5em minus 0.4em Philadelphia, PA: SIAM, 2003, vol. 45

    2003

  50. [50]

    T. Yu, S. Kumar, A. Gupta, S. Levine, K. Hausman, and C. Finn, ``Gradient surgery for multi-task learning,'' Adv. Neural Inf. Process. Syst., vol. 33, pp. 5824--5836, 2020

    2020

  51. [51]

    Jaegle, F

    A. Jaegle, F. Gimeno, A. Brock, O. Vinyals, A. Zisserman, and J. Carreira, ``Perceiver: General perception with iterative attention,'' in International Conference on Machine Learning (ICML), 2021, pp. 4651--4664

    2021

  52. [52]

    Carion, F

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, ``End-to-end object detection with transformers,'' in European Conference on Computer Vision (ECCV), 2020

    2020

  53. [53]

    Bourtsoulatze, D

    E. Bourtsoulatze, D. B. Kurka, and D. G \"u nd \"u z, ``Deep joint source-channel coding for wireless image transmission,'' IEEE Trans. Cogn. Commun. Netw., vol. 5, no. 3, pp. 567--579, 2019

    2019

  54. [54]

    J. Shao, Y. Mao, and J. Zhang, ``Learning task-oriented communication for edge inference: An information bottleneck approach,'' IEEE J. Sel. Areas Commun., vol. 40, no. 1, pp. 197--211, 2022

    2022

  55. [55]

    K. Cho et al., ``Learning phrase representations using RNN encoder--decoder for statistical machine translation,'' in Conference on Empirical Methods in Natural Language Processing (EMNLP), 2014, pp. 1724--1734

    2014

  56. [56]

    Z. Liu, H. Mao, C.-Y. Wu, C. Feichtenhofer, T. Darrell, and S. Xie, ``A ConvNet for the 2020s,'' in IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 11\,966--11\,976

    2022

  57. [57]

    Deng, ``The MNIST database of handwritten digit images for machine learning research [best of the web],'' IEEE Signal Process

    L. Deng, ``The MNIST database of handwritten digit images for machine learning research [best of the web],'' IEEE Signal Process. Mag., vol. 29, no. 6, pp. 141--142, 2012

    2012

  58. [58]

    H. Xiao, K. Rasul, and R. Vollgraf, `` Fashion-MNIST : a novel image dataset for benchmarking machine learning algorithms,'' arXiv preprint arXiv:1708.07747, 2017

    Pith/arXiv arXiv 2017

  59. [59]

    Z. Liu, P. Luo, X. Wang, and X. Tang, ``Deep learning face attributes in the wild,'' in IEEE International Conference on Computer Vision (ICCV), December 2015

    2015

  60. [60]

    Shaler, DanGill, Maggie, M

    B. Shaler, DanGill, Maggie, M. McDonald, Patricia, and W. Cukierski, ``Carvana image masking challenge,'' https://kaggle.com/competitions/carvana-image-masking-challenge, 2017

    2017

  61. [61]

    Zheng, Y

    X. Zheng, Y. Wang, G. Wang, and J. Liu, ``Fast and robust segmentation of white blood cell images by self-supervised learning,'' Micron, vol. 107, pp. 55--71, 2018

    2018

  62. [62]

    C. Dong, C. C. Loy, and X. Tang, ``Accelerating the super-resolution convolutional neural network,'' in European Conference on Computer Vision (ECCV), 2016, pp. 391--407

    2016

  63. [63]

    W. Shi et al., ``Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network,'' in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 1874--1883

    2016

  64. [64]

    Mildenhall, P

    B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoorthi, and R. Ng, `` NeRF : Representing scenes as neural radiance fields for view synthesis,'' in European Conference on Computer Vision (ECCV), 2020

    2020

  65. [65]

    Liu et al., `` KAN : Kolmogorov--Arnold networks,'' arXiv preprint arXiv:2404.19756, 2024

    Z. Liu et al., `` KAN : Kolmogorov--Arnold networks,'' arXiv preprint arXiv:2404.19756, 2024

    Pith/arXiv arXiv 2024

  66. [66]

    Gu and T

    A. Gu and T. Dao, ``Mamba: Linear-time sequence modeling with selective state spaces,'' in Conference on Language Modeling (COLM), 2024

    2024

  67. [67]

    Cheng, R

    B. Cheng, R. Girshick, P. Doll \'a r, A. C. Berg, and A. Kirillov, ``Boundary IoU : Improving object-centric image segmentation evaluation,'' in IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 15\,334--15\,342

    2021

  68. [68]

    Maier-Hein, A

    L. Maier-Hein, A. Reinke, P. Godau, M. Tizabi, F. Buettner et al., ``Metrics reloaded: recommendations for image analysis validation,'' Nat. Methods, vol. 21, no. 2, pp. 195--212, 2024

    2024