Pith. sign in

REVIEW 2 major objections 48 references

FoundDP restores structural consistency in dual-pixel depth by aligning monocular ViT features to offset defocus blur while retaining metric scale from the disparity signal.

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 · grok-4.3

2026-07-03 16:05 UTC pith:J6JQB5K2

load-bearing objection FoundDP adds a ViT feature alignment step to handle weak disparity in dual-pixel depth, and the construction holds up without obvious flaws. the 2 major comments →

arxiv 2607.01900 v1 pith:J6JQB5K2 submitted 2026-07-02 cs.CV

FoundDP: Revisiting Weak Disparity Observability in Dual-Pixel Depth Estimation

Fengchen He , Hao Xu , Dayang Zhao , Tingwei Quan , Shaoqun Zeng This is my paper
classification cs.CV
keywords dual-pixel depth estimationweak disparity observabilitymonocular depth priorsVision Transformer feature alignmentdefocus blur mitigationmetric depth from single camera
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.

Dual-pixel sensors extract metric depth from sub-aperture disparity, yet the short baseline makes disparity signals weak or absent in textureless, low-contrast, or downsampled areas, so local-cue methods produce broken structures and inaccurate depths. The paper establishes that these failures stem from over-reliance on ambiguous local disparity and shows they can be corrected by importing global structural priors from a monocular depth foundation model. To make the transfer reliable under DP imaging, the method identifies how defocus blur degrades ViT representations and counters it with a dedicated feature-alignment step that keeps the monocular priors stable. The resulting estimates therefore combine the metric fidelity of the DP signal with the structural completeness supplied by the foundation model, and this advantage grows precisely where disparity observability is lowest. Experiments on both synthetic and real DP datasets confirm consistent gains in structural fidelity and metric accuracy under those conditions.

Core claim

The central claim is that ViT feature alignment can mitigate representation degradation caused by DP defocus blur, thereby enabling stable integration of metric depth from dual-pixel disparity with global structural priors from a monocular foundation model so that depth estimation remains accurate even when local disparity cues are weak or ambiguous.

What carries the argument

The ViT feature alignment step that counters DP-induced representation degradation so monocular structural priors can be fused reliably with metric DP depth.

Load-bearing premise

That feature alignment can neutralize DP defocus blur effects on the monocular model without creating new scale or structural inconsistencies.

What would settle it

On real DP test images with known weak-disparity regions, run the alignment step and measure whether structural fidelity or metric accuracy still drops relative to the unaligned monocular baseline.

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

If this is right

  • Depth maps keep the absolute scale supplied by the dual-pixel baseline while acquiring the missing structure that local disparity cues cannot provide.
  • Gains appear most clearly on textureless and downsampled regions where conventional DP methods degrade.
  • The same alignment mechanism stabilizes the transfer of any monocular foundation model to DP data without retraining the entire model.

Where Pith is reading between the lines

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

  • The same alignment idea could be tested on other single-camera depth modalities that suffer from blur-induced feature shifts.
  • If alignment proves general, it may reduce the need for task-specific fine-tuning when moving foundation models between different camera optics.
  • A direct comparison of alignment cost versus performance on progressively smaller baselines would quantify how far the approach extends the usable range of DP sensors.

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

2 major / 0 minor

Summary. The manuscript proposes FoundDP, a framework for dual-pixel (DP) depth estimation that integrates metric depth from the DP branch with global structural priors from a monocular ViT-based foundation model. It identifies ViT feature degradation caused by DP defocus blur and mitigates it through a feature alignment step, with the goal of restoring structural consistency in regions of weak disparity observability while preserving metric scale from the DP supervision. The abstract claims that extensive experiments on synthetic and real-world DP benchmarks demonstrate superior performance in structural fidelity and metric accuracy, particularly under reduced disparity observability.

Significance. If the quantitative results, ablations, and derivations support the claims, the work would offer a targeted way to combine the metric accuracy of DP imaging with the structural robustness of foundation models, addressing a practical limitation in single-camera depth estimation. The focus on weak-disparity regions and the explicit handling of blur-induced ViT degradation represent a relevant technical contribution to the DP depth literature.

major comments (2)
  1. [Abstract] Abstract: The abstract asserts 'superior performance, with consistent gains in structural fidelity and metric accuracy' and 'extensive experiments' but supplies no quantitative metrics, benchmark names with scores, ablation results, or derivation of the feature alignment procedure. This absence makes the central performance claims unverifiable from the provided text and prevents assessment of whether the alignment step introduces scale inconsistencies or structural artifacts.
  2. [Abstract] The description of the ViT feature alignment step lacks any equation, loss formulation, or pseudocode. Without these details it is impossible to evaluate whether the alignment is applied only to weak-disparity regions, whether it is parameter-free, or how it interacts with the DP depth head that is stated to be the sole source of metric supervision.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The abstract asserts 'superior performance, with consistent gains in structural fidelity and metric accuracy' and 'extensive experiments' but supplies no quantitative metrics, benchmark names with scores, ablation results, or derivation of the feature alignment procedure. This absence makes the central performance claims unverifiable from the provided text and prevents assessment of whether the alignment step introduces scale inconsistencies or structural artifacts.

    Authors: We agree that the abstract, as a concise summary, omits specific numerical results and derivations. To improve verifiability of the claims, we will revise the abstract to include key quantitative metrics (e.g., MAE and structural similarity gains on the synthetic and real-world DP benchmarks) while keeping it brief. Full ablations, derivations, and analysis confirming that the alignment preserves metric scale without introducing artifacts are already in Sections 4 and 5. revision: yes

  2. Referee: [Abstract] The description of the ViT feature alignment step lacks any equation, loss formulation, or pseudocode. Without these details it is impossible to evaluate whether the alignment is applied only to weak-disparity regions, whether it is parameter-free, or how it interacts with the DP depth head that is stated to be the sole source of metric supervision.

    Authors: The abstract provides only a high-level overview. The equations, loss formulation, selective application to weak-disparity regions, parameter efficiency, and interaction with the DP depth head (sole metric supervisor) are fully specified in Section 3.2. We will revise the abstract to briefly note that the alignment is a targeted, parameter-efficient module preserving DP metric scale, directing readers to the main text for details. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper proposes FoundDP as a framework combining metric DP depth with monocular foundation model priors via ViT feature alignment to handle weak disparity. No equations, fitting procedures, or self-referential definitions appear in the abstract or description that would reduce any claimed prediction or result to its inputs by construction. Performance is asserted via experiments on external synthetic and real-world DP benchmarks. No load-bearing self-citations, uniqueness theorems, or ansatzes imported from prior author work are referenced. The derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides no equations, training details, or modeling choices; no free parameters, axioms, or invented entities can be identified.

pith-pipeline@v0.9.1-grok · 5737 in / 1027 out tokens · 20751 ms · 2026-07-03T16:05:56.949644+00:00 · methodology

0 comments
read the original abstract

Dual-pixel (DP) imaging enables metric depth estimation from a single camera using sub-aperture disparity. However, the extremely small effective baseline limits disparity observability, leading to structural degradation and depth failure in textureless, low-contrast, or downsampled regions. Existing DP-based methods rely primarily on local disparity cues and therefore become unreliable when disparity signals are weak or ambiguous. To address this limitation, we propose \emph{FoundDP}, a unified framework that integrates metric DP depth with global structural priors from a monocular depth foundation model. Our method preserves metric scale through DP-derived depth and leverages Vision Transformer (ViT) features to restore structural consistency in weak-disparity regions. To ensure reliable metric guidance under DP imaging conditions, we identify and mitigate ViT representation degradation induced by DP defocus blur via ViT feature alignment, enabling stable metric-guided depth estimation. Extensive experiments on synthetic and real-world DP benchmarks show that FoundDP delivers superior performance, with consistent gains in structural fidelity and metric accuracy, especially under reduced disparity observability. Code will be available at: https://github.com/EchoLighting/FoundDP

Figures

Figures reproduced from arXiv: 2607.01900 by Dayang Zhao, Fengchen He, Hao Xu, Shaoqun Zeng, Tingwei Quan.

Figure 1
Figure 1. Figure 1: (a) Qualitative comparison. Our method yields sharper boundaries and more ac￾curate depth in weak-disparity regions (highlighted in red). (b) Complementary failure modes: DPNet preserves metric scale but degrades structurally, while DAV2 recovers structure but lacks metric scale. Best viewed by zooming in. In contrast, recent monocular depth foundation models based on Vision Transformers (ViT) [24, 27, 34,… view at source ↗
Figure 2
Figure 2. Figure 2: (a) Overall framework of FoundDP. DP sub-aperture images are used to esti￾mate metric depth Ddp, which is refined to Dmetric for improved geometric consistency. In parallel, a monocular depth foundation model extracts structural priors. A depth guidance module then uses metric depth to guide the structural features and produce the final depth Dguide. (b) Illustration of depth histogram analysis. (1) Compar… view at source ↗
Figure 3
Figure 3. Figure 3: Effect of ViT alignment under defo￾cus degradation. Aligned ViT features yield more structurally consistent predictions in blurred regions. To mitigate this degradation, we apply explicit ViT feature alignment prior to restoring structural consis￾tency. Specifically, paired clear RGB images and their DP-degraded coun￾terparts are processed by a shared ViT encoder, and a feature-space alignment constraint i… view at source ↗
Figure 4
Figure 4. Figure 4: Qualitative comparison. Rows 1-3: Results on representative DP datasets. Our method preserves sharper structural boundaries and more coherent geometry, es￾pecially in weak-disparity regions (highlighted in red). Rows 4-5: Results on the down￾sampling dataset. Our model remains robust under disparity attenuation, whereas com￾peting methods exhibit structural degradation. 4.2 Datasets and Evaluation Metrics … view at source ↗
Figure 5
Figure 5. Figure 5: Depth accuracy versus dispar￾ity observability. Our method shows im￾proved robustness under reduced observ￾ability [PITH_FULL_IMAGE:figures/full_fig_p013_5.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

48 extracted references · 48 canonical work pages · 1 internal anchor

  1. [1]

    In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision

    Abuolaim, A., Afifi, M., Brown, M.S.: Improving single-image defocus deblurring: How dual-pixel images help through multi-task learning. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. pp. 1231– 1239 (2022)

    work page 2022

  2. [2]

    Abuolaim,A.,Brown,M.S.:Defocusdeblurringusingdual-pixeldata.In:European conference on computer vision. pp. 111–126. Springer (2020)

    work page 2020

  3. [3]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops

    Bastidas, A.A., Tang, H.: Channel attention networks. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops. pp. 0–0 (2019)

    work page 2019

  4. [4]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Choi, M., Lee, H., Lee, H.e.: Exploring positional characteristics of dual-pixel data for camera autofocus. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 13158–13168 (2023)

    work page 2023

  5. [5]

    Doehyung, L., Zhuofeng, W., Yusuke, M., Masatoshi, O.: Fmdp: Leveraging a foun- dationmodelfordual-pixeldisparityestimation.IEICEProceedingSeries93,O1–1 (2025)

    work page 2025

  6. [6]

    An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

    Dosovitskiy, A.: An image is worth 16x16 words: Transformers for image recogni- tion at scale. arXiv preprint arXiv:2010.11929 (2020)

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  7. [7]

    Advances in neural information processing systems27 (2014)

    Eigen, D., Puhrsch, C., Fergus, R.: Depth map prediction from a single image using a multi-scale deep network. Advances in neural information processing systems27 (2014)

    work page 2014

  8. [8]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Fu, H., Gong, M., Wang, C., Batmanghelich, K., Tao, D.: Deep ordinal regression network for monocular depth estimation. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2002–2011 (2018)

    work page 2002

  9. [9]

    In: Proceedings of the IEEE/CVF international con- ference on computer vision

    Garg, R., Wadhwa, N., Ansari, S., Barron, J.T.: Learning single camera depth estimation using dual-pixels. In: Proceedings of the IEEE/CVF international con- ference on computer vision. pp. 7628–7637 (2019)

    work page 2019

  10. [10]

    In: ProceedingsoftheIEEE/CVFConferenceonComputerVisionandPatternRecog- nition

    Ghanekar, B., Khan, S.S., Sharma, P., Singh, S., Boominathan, V., Mitra, K., Veeraraghavan, A.: Passive snapshot coded aperture dual-pixel rgb-d imaging. In: ProceedingsoftheIEEE/CVFConferenceonComputerVisionandPatternRecog- nition. pp. 25348–25357 (2024)

    work page 2024

  11. [11]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Godard, C., Mac Aodha, O., Brostow, G.J.: Unsupervised monocular depth es- timation with left-right consistency. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 270–279 (2017)

    work page 2017

  12. [12]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    He, F., Zhao, D., Xu, H., Quan, T., Zeng, S.: Simulating dual-pixel images from ray tracing for depth estimation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 26106–26115 (2025)

    work page 2025

  13. [13]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Jiang, H., Lou, Z., Ding, L., Xu, R., Tan, M., Jiang, W., Huang, R.: Defom-stereo: Depth foundation model based stereo matching. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 21857–21867 (2025) 16 F. He et al

    work page 2025

  14. [14]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recog- nition

    Ke, B., Obukhov, A., Huang, S., Metzger, N., Daudt, R.C., Schindler, K.: Re- purposing diffusion-based image generators for monocular depth estimation. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recog- nition. pp. 9492–9502 (2024)

    work page 2024

  15. [15]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Kim, D., Jang, H., Kim, I., Kim, M.H.: Spatio-focal bidirectional disparity esti- mation from a dual-pixel image. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 5023–5032 (2023)

    work page 2023

  16. [16]

    ITE Transactions on Media Technology and Applications4(2), 123–128 (2016)

    Kobayashi, M., Johnson, M., Wada, Y., Tsuboi, H., Takada, H., Togo, K., Kishi, T., Takahashi,H.,Ichikawa,T.,Inoue,S.:Alownoiseandhighsensitivityimagesensor with imaging and phase-difference detection af in all pixels. ITE Transactions on Media Technology and Applications4(2), 123–128 (2016)

    work page 2016

  17. [17]

    In: 2023 IEEE International Con- ference on Computational Photography (ICCP)

    Li, F., Guo, H., Santo, H., Okura, F., Matsushita, Y.: Learning to synthesize pho- torealistic dual-pixel images from rgbd frames. In: 2023 IEEE International Con- ference on Computational Photography (ICCP). pp. 1–11. IEEE (2023)

    work page 2023

  18. [18]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    Li, Y., Monno, Y., Okutomi, M.: Dual-pixel raindrop removal. IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    work page 2024

  19. [19]

    Neurocomputing616, 128880 (2025)

    Li, Y., Yi, Y., Shu, X., Ren, D., Li, Q., Zuo, W.: Learning dual-pixel alignment for defocus deblurring. Neurocomputing616, 128880 (2025)

    work page 2025

  20. [20]

    In: Proceedings of the Computer Vision and Pattern Recognition Con- ference

    Lin, H., Peng, S., Chen, J., Peng, S., Sun, J., Liu, M., Bao, H., Feng, J., Zhou, X., Kang, B.: Prompting depth anything for 4k resolution accurate metric depth estimation. In: Proceedings of the Computer Vision and Pattern Recognition Con- ference. pp. 17070–17080 (2025)

    work page 2025

  21. [21]

    In: 2024 International Conference on 3D Vision (3DV)

    Monin, S., Katz, S., Evangelidis, G.: Continuous cost aggregation for dual-pixel disparity extraction. In: 2024 International Conference on 3D Vision (3DV). pp. 675–684. IEEE (2024)

    work page 2024

  22. [22]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Pan, L., Chowdhury, S., Hartley, R., Liu, M., Zhang, H., Li, H.: Dual pixel ex- ploration: Simultaneous depth estimation and image restoration. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 4340–4349 (2021)

    work page 2021

  23. [23]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    Pan, L., Hartley, R., Liu, L., Xu, Z., Chowdhury, S., Yang, Y., Zhang, H., Li, H., Liu, M.: Weakly-supervised depth estimation and image deblurring via dual-pixel sensors. IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    work page 2024

  24. [24]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Piccinelli, L., Yang, Y.H., Sakaridis, C., Segu, M., Li, S., Van Gool, L., Yu, F.: Unidepth: Universal monocular metric depth estimation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 10106– 10116 (2024)

    work page 2024

  25. [25]

    In: 2020 IEEE International Conference on Com- putational Photography (ICCP)

    Punnappurath, A., Abuolaim, A., Afifi, M., Brown, M.S.: Modeling defocus- disparity in dual-pixel sensors. In: 2020 IEEE International Conference on Com- putational Photography (ICCP). pp. 1–12. IEEE (2020)

    work page 2020

  26. [26]

    In: ProceedingsoftheIEEE/CVFConferenceonComputerVisionandPatternRecog- nition

    Punnappurath, A., Brown, M.S.: Reflection removal using a dual-pixel sensor. In: ProceedingsoftheIEEE/CVFConferenceonComputerVisionandPatternRecog- nition. pp. 1556–1565 (2019)

    work page 2019

  27. [27]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Ranftl, R., Bochkovskiy, A., Koltun, V.: Vision transformers for dense prediction. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 12179–12188 (2021)

    work page 2021

  28. [28]

    IEEE transactions on pattern analysis and machine intelligence44(3), 1623–1637 (2020) FoundDP 17

    Ranftl, R., Lasinger, K., Hafner, D., Schindler, K., Koltun, V.: Towards robust monocular depth estimation: Mixing datasets for zero-shot cross-dataset transfer. IEEE transactions on pattern analysis and machine intelligence44(3), 1623–1637 (2020) FoundDP 17

    work page 2020

  29. [29]

    ACM Transactions on Graphics (TOG)43(4), 1–19 (2024)

    Shi, Z., Chugunov, I., Bijelic, M., Côté, G., Yeom, J., Fu, Q., Amata, H., Heidrich, W., Heide, F.: Split-aperture 2-in-1 computational cameras. ACM Transactions on Graphics (TOG)43(4), 1–19 (2024)

    work page 2024

  30. [30]

    In: European conference on computer vision

    Silberman, N., Hoiem, D., Kohli, P., Fergus, R.: Indoor segmentation and support inference from rgbd images. In: European conference on computer vision. pp. 746–

    work page

  31. [31]

    Journal of Computer and Communications1(06), 11 (2013)

    Śliwiński, P., Wachel, P.: A simple model for on-sensor phase-detection autofocus- ing algorithm. Journal of Computer and Communications1(06), 11 (2013)

    work page 2013

  32. [32]

    ACM Transactions on Graphics (ToG)37(4), 1–13 (2018)

    Wadhwa, N., Garg, R., Jacobs, D.E., Feldman, B.E., Kanazawa, N., Carroll, R., Movshovitz-Attias, Y., Barron, J.T., Pritch, Y., Levoy, M.: Synthetic depth-of-field with a single-camera mobile phone. ACM Transactions on Graphics (ToG)37(4), 1–13 (2018)

    work page 2018

  33. [33]

    Wang, Q., Wu, B., Zhu, P., Li, P., Zuo, W., Hu, Q.: Eca-net: Efficient channel attentionfordeepconvolutionalneuralnetworks.In:ProceedingsoftheIEEE/CVF conference on computer vision and pattern recognition. pp. 11534–11542 (2020)

    work page 2020

  34. [34]

    In: Proceedings of the Computer Vision and Pattern Recog- nition Conference

    Wang, R., Xu, S., Dai, C., Xiang, J., Deng, Y., Tong, X., Yang, J.: Moge: Unlock- ing accurate monocular geometry estimation for open-domain images with optimal training supervision. In: Proceedings of the Computer Vision and Pattern Recog- nition Conference. pp. 5261–5271 (2025)

    work page 2025

  35. [35]

    Depth Anything with Any Prior

    Wang, Z., Chen, S., Yang, L., Wang, J., Zhang, Z., Zhao, H., Zhao, Z.: Depth anything with any prior. arXiv preprint arXiv:2505.10565 (2025)

    work page Pith review arXiv 2025

  36. [36]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Wen, B., Trepte, M., Aribido, J., Kautz, J., Gallo, O., Birchfield, S.: Foundation- stereo: Zero-shot stereo matching. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 5249–5260 (2025)

    work page 2025

  37. [37]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Wu, K., Peng, H., Chen, M., Fu, J., Chao, H.: Rethinking and improving rela- tive position encoding for vision transformer. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 10033–10041 (2021)

    work page 2021

  38. [38]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Xin, S., Wadhwa, N., Xue, T., Barron, J.T., Srinivasan, P.P., Chen, J., Gkioulekas, I., Garg, R.: Defocus map estimation and deblurring from a single dual-pixel image. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 2228–2238 (2021)

    work page 2021

  39. [39]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Xu, D., Ouyang, W., Wang, X., Sebe, N.: Pad-net: Multi-tasks guided prediction- and-distillation network for simultaneous depth estimation and scene parsing. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 675–684 (2018)

    work page 2018

  40. [40]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Yang, H., Pan, L., Yang, Y., Hartley, R., Liu, M.: Ldp: Language-driven dual-pixel image defocus deblurring network. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 24078–24087 (2024)

    work page 2024

  41. [41]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Yang, L., Kang, B., Huang, Z., Xu, X., Feng, J., Zhao, H.: Depth anything: Un- leashing the power of large-scale unlabeled data. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 10371–10381 (2024)

    work page 2024

  42. [42]

    Advances in Neural Information Processing Systems37, 21875–21911 (2024)

    Yang, L., Kang, B., Huang, Z., Zhao, Z., Xu, X., Feng, J., Zhao, H.: Depth anything v2. Advances in Neural Information Processing Systems37, 21875–21911 (2024)

    work page 2024

  43. [43]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Yang, Y., Pan, L., Liu, L., Liu, M.: K3dn: Disparity-aware kernel estimation for dual-pixel defocus deblurring. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 13263–13272 (2023)

    work page 2023

  44. [44]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Ye, T., Chen, S., Chen, H., Chai, W., Ren, J., Xing, Z., Li, W., Zhu, L.: Prompt- haze: Prompting real-world dehazing via depth anything model. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 39, pp. 9454–9462 (2025) 18 F. He et al

    work page 2025

  45. [45]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Yin, W., Zhang, J., Wang, O., Niklaus, S., Mai, L., Chen, S., Shen, C.: Learning to recover 3d scene shape from a single image. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 204–213 (2021)

    work page 2021

  46. [46]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference

    Yu, H., Song, S., Sun, L., Su, W., Yang, X., Liu, C.: All-directional disparity estimation for real-world qpd images. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 21836–21846 (2025)

    work page 2025

  47. [47]

    In: Proceedings of the 33rd ACM International Conference on Multimedia

    Yu, K., Pan, L., Liu, L., Liang, W.: Enhanced dual-pixel image reflection removal via gaussian splatting. In: Proceedings of the 33rd ACM International Conference on Multimedia. pp. 7766–7775 (2025)

    work page 2025

  48. [48]

    In: European Con- ference on Computer Vision

    Zhang, Y., Wadhwa, N., Orts-Escolano, S., Häne, C., Fanello, S., Garg, R.: Du2net: Learning depth estimation from dual-cameras and dual-pixels. In: European Con- ference on Computer Vision. pp. 582–598. Springer (2020)

    work page 2020