Pith. sign in

REVIEW 3 major objections 6 minor 98 references

DMS:Diffusion-Based Multi-Baseline Stereo Generation for Improving Self-Supervised Depth Estimation

T0 review · 3 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A diffusion model that synthesizes extra stereo views can cut disparity outliers by up to 35 percent in self-supervised depth estimation.

desk verdict A novel diffusion-based multi-baseline view generator that improves self-supervised depth, but the center-view geometry rests on an unverified upscaling assumption and needs a direct disparity check. read the letter →

arxiv 2508.13091 v1 pith:UOTGTXDB submitted 2025-08-18 cs.CV

Zihua Liu , Yizhou Li , Songyan Zhang , Masatoshi Okutomi This is my paper · ORCID
classification cs.CV
keywords self-superviseddepthestimationstereomatchingdiffusionmodelsnovelviewsynthesismulti-baselinephotometricwarpinglossStablemonocular
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

Self-supervised depth estimators trained on stereo pairs struggle where pixels are visible in only one camera — occlusions and out-of-frame areas — because the photometric reconstruction loss has no matching pixel to use. The paper proposes DMS, a plug-in that fine-tunes a latent diffusion model to generate additional views along the epipolar line: a left-left view, a right-right view, and a center view between the two cameras. These synthesized views supply explicit correspondences, so a per-pixel minimum warping loss can pick the source view that actually sees each pixel. Training and synthesis need only unlabeled stereo pairs, and the authors report up to 35% outlier reduction and improved disparity maps across several benchmarks.

What carries the argument

The mechanism is the repurposed latent diffusion model used as an epipolar novel-view synthesizer: it encodes the reference image into a latent, concatenates the target latent during training, and is steered at inference by text direction prompts ('to left' / 'to right') to produce left-left and right-right views. An intermediate view is obtained by a resolution-scale trick — upscaling the input by X=2 before generation produces the same pixel displacement in the upscaled coordinates, hence half the disparity at the original resolution. The synthesized views enter the Per-Pixel Minimum Warping Loss, which takes the minimum photometric error over source views {right, left-left, right-right, center} with disparity scale factors {1, -1, 2, 0.5}, so each pixel is matched from whichever generated view actually sees it.

What would settle it

In a CARLA scene with ground-truth multi-baseline views, generate the center view from the left image at X=2 and compute the end-point error between this generated view warped by true disparity and the true center image; if the center view does not sit at half the left-right disparity, the geometric-consistency assumption behind the minimum warping loss fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that a latent diffusion model fine-tuned on ordinary unlabeled stereo pairs can act as a multi-baseline stereo generator: conditioned on one image and a direction prompt ('to left' or 'to right'), it produces geometrically consistent novel views beyond the original pair, and upscaling the input by a factor X=2 yields an intermediate center view with proportionally smaller horizontal displacement. When these views are fed into the per-pixel minimum warping loss alongside the original right image, occluded and out-of-frame regions gain explicit photometric correspondences, improving disparity estimation for self-supervised stereo matching and stereo-supervised monocular depth estimation. The method is model-agnostic and adds no labeled supervision.

Load-bearing premise

The load-bearing premise is that upscaling the diffusion model's input by a factor X shrinks the generated horizontal displacement by the same factor, so the X=2 output really is a center view; the paper treats this as an empirical finding and supports it only indirectly with PSNR/SSIM on CARLA, not by measuring the generated center view's disparity.

Editorial extensions

If this is right

  • Adding the generated left-left, right-right, and center views reduces outlier ratios substantially in occluded and out-of-frame regions, for example from 83.6% to 49.2% occlusion outliers on SceneFlow with the baseline stereo network.
  • DMS improves both classical stereo (SGM) and learning-based stereo networks on the KITTI 2015 benchmark without changing their architectures.
  • The same generated views improve stereo-supervised monocular depth estimators on the KITTI Eigen split.
  • Combining all generated views outperforms any single one; the center view alone gives the best intermediate result, and adding it to left-left and right-right yields the largest gains.
  • Because the method uses only unlabeled stereo pairs for both synthesis and training, it can be applied to any new stereo dataset without extra annotations.

Reading between the lines

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

  • If the resolution-scaling observation generalizes to arbitrary factors, the same diffusion generator could produce a continuum of baselines rather than fixed units, enabling adaptive baseline selection for different depth ranges; the paper only tests X=2 as the center view.
  • The same multi-view minimum-warping supervision should transfer to self-supervised optical flow or scene flow, where occlusion creates the identical missing-correspondence problem; the paper does not test those tasks.
  • A stricter check of geometric consistency would measure the generated center view's disparity against ground truth (e.g., on a multi-baseline simulator) instead of PSNR/SSIM; if the X=2 view deviates from half disparity, photometric matching would bias depth estimates at object boundaries.
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.

Referee Report

3 major / 6 minor

Summary. The paper proposes DMS, a two-stage method for self-supervised depth estimation. In stage 1, a Stable Diffusion V2 model is fine-tuned on unlabeled stereo pairs with direction prompts ('to left'/'to right') to synthesize multi-baseline views: left-left, right-right, and a center view obtained by upscaling the input by a factor X before generation. In stage 2, these views are incorporated into the photometric training of self-supervised stereo matching and stereo-supervised monocular depth networks through a per-pixel minimum warping loss (Eq. 5) with fixed disparity scale factors. Experiments on SceneFlow, KITTI, and MPI-Sintel show consistent EPE/D1 improvements, including up to about 35% outlier reduction on the KITTI 2015 test set, and the method is demonstrated across several stereo and monocular backbones.

Significance. If the geometric consistency of the generated views were established, DMS would be a practically valuable and unusually simple contribution: it turns a large pretrained diffusion model into a multi-baseline stereo generator without extra annotations, and it improves several self-supervised depth baselines in a model-agnostic way. The paper's strengths include the large-scale evaluation across datasets and network architectures, the inclusion of ablations that separate the contributions of left-left/right-right and center views, and the public code release. The main open risk is that the central geometric claim is asserted rather than verified; the currently reported appearance metrics are not sufficient to support it.

major comments (3)
  1. [Section 3.2.3 and Eq. (5)] The generation of the center view rests on an unverified scale-invariance assumption. The text states that "upscaling the input resolution yields the same pixel-level displacements in generated views," but the paper provides neither a derivation nor a direct measurement of the generated disparity. The CARLA evaluation in Table 10 reports only PSNR and SSIM, which do not establish that the center view's disparity is d/2 or that the left-left and right-right views have disparities of -d and 2d. Since Eq. (5) uses the fixed scales s_c=0.5, s_ll=-1, and s_rr=2, an incorrect scale factor would allow the min operator to select the view whose artifacts best match the target and still reduce the photometric loss, so the EPE/D1 gains in Tables 5 and 6 would not demonstrate correct geometric supervision. Please add a direct disparity-level evaluation on CARLA (e.g., EPE/D1 of the generated views against the simulator ground truth, or per-view warping error under GT disparity) and report the actual measured displacement scale for each generated view.
  2. [Section 4.3.2 and Table 6] The benchmark comparison may not isolate the effect of DMS. For the learning-based entries marked with an asterisk, both the baseline and +DMS numbers are re-implemented "after unsupervised training with vanilla warping loss," but the paper does not state whether the two rows use identical training budgets, augmentations, hyperparameters, and random seeds, nor does it report variance or standard error for the KITTI test results. Since several methods show more than 30% D1-bg reduction, a small difference in the fine-tuning protocol could explain part of the gain. Please specify the exact protocol and, ideally, report the re-implemented baseline alongside the original published numbers so the reader can verify that the comparison is controlled.
  3. [Supplementary Section 5.1.3 and Table 5] The occluded-region evaluation on KITTI relies on pseudo-occlusion masks produced by a pre-trained stereo model rather than ground-truth occlusions, while the text says "ground disparity" is used for the mask formula. This makes the reported "improvement in occluded regions" on KITTI harder to interpret, as errors in the pseudo masks are correlated with the quality of the stereo method being evaluated. Please clarify the exact mask source for each dataset in Table 5 and, if possible, report the occluded-region numbers on SceneFlow and MPI-Sintel, where ground-truth occlusion masks are available, as the primary evidence.
minor comments (6)
  1. [Section 3.3, Eq. (6)] The weighting parameter alpha in the photometric loss is never defined in the main text; please state its value or reference where it is set.
  2. [Section 4.3.2] The sentence referring to "Equation 3.3" is ambiguous; it should reference the per-pixel minimum warping loss (Eq. 5) or the exact equation used for the SGM cost volume construction.
  3. [Section 4.1 and Table 8] The abstract and introduction call DMS "cost-free," but Table 8 reports 4-6 seconds per image and about 7 GB GPU memory for inference; please qualify the term, for example by stating that no additional labels or network modifications are required.
  4. [Supplementary Section 5.1.3] There is a typo in the mask generation description: "pseudo-left and pseudo-left disparities" should presumably read "pseudo-left and pseudo-right disparities"; please also clarify whether the KITTI occlusion masks used in Table 5 come from this pre-trained model and how its accuracy affects the reported occluded-region gains.
  5. [Various] Several typos should be corrected: "ScenceFlow" in Section 4.3.1 and Table 5, "Gemetric Consistency" in Table 3, and "Denosing UNet" in Section 4.1.
  6. [Table 10 caption] The phrase "check-marked annotations" is unclear; state explicitly which rows are novel views generated by the proposed inference procedure rather than the existing left/right views.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: DMS adds independently generated synthetic views to a photometric loss, and improvements are measured on external ground-truth benchmarks.

full rationale

The paper's central chain is: (1) fine-tune a Stable Diffusion model on unlabeled stereo pairs to synthesize opposite views; (2) at inference, generate left-left, right-right, and center views by direction prompts and resolution rescaling; (3) use these views as additional sources in the per-pixel minimum warping loss (Eq. 5); (4) evaluate the resulting disparity/depth on SceneFlow, KITTI, and MPI-Sintel test sets with ground truth. No step reduces by construction to its own inputs. The generated views are fixed synthetic images produced before depth training, not fitted to the depth network's predictions, so the 'prediction' is not a renamed fit. The scale factors s_i in Eq. 5 are set by the geometric baseline of each view; the center-view scale s_c = 0.5 is tied to the upscaling factor X = 2.0 through the empirical observation of pixel-displacement invariance under rescaling. That is an unverified empirical premise (and therefore a correctness risk, as the paper only reports PSNR/SSIM on CARLA rather than direct disparity checks), but it is not a definitional or self-referential equivalence. The self-citations [45,46] appear in related work and as a mask-generation recipe in the supplement; they are not load-bearing for the central DMS claim. The ablation and benchmark tables compare the same network with and without DMS views on external ground truth, so the reported gains are not forced by construction.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on two empirical assumptions about the diffusion model's behavior: that directional prompts produce views with the intended epipolar shift, and that upscaling creates a geometrically valid intermediate view. Neither is derived or directly measured. The scale factors in the loss are assumed constants. There are no new physical entities.

free parameters (2)
  • Rescale factor X for intermediate view = 2.0
    Chosen by hand and supported by a small ablation (Table 9 in the supplementary) showing that X=2.0 gives near-best performance. It is not derived from first principles and is a tunable knob.
  • Scale factors in warping loss (sr, sll, srr, sc) = 1, -1, 2, 0.5
    These assume the generated views have exact baseline multiples relative to the left view. They are not measured from the outputs; they are assumed constants in Equation (5). If the diffusion model does not produce exactly these shifts, the loss is biased.
assumptions (4)
  • domain assumption The fine-tuned Stable Diffusion model generates views with the expected epipolar shift when prompted with 'to left' or 'to right'.
    The method relies on the model having implicitly learned the stereo disparity distribution from the training pairs. This is stated in Section 3.2 and is not formally proven.
  • ad hoc to paper Upscaling the input image by X produces an intermediate view with displacement scaled by 1/X.
    Introduced in Section 3.2.3 as an empirical finding ('we found that upscaling the input resolution yields the same pixel-level displacements in generated views'). It is the central geometric trick for center-view generation and is only qualitatively verified.
  • domain assumption The generated views are geometrically consistent enough to serve as photometric supervision for depth learning.
    The per-pixel minimum warping loss treats the synthetic views as if they were real images from additional cameras. The paper provides indirect evidence (SIFT matches, warping visualization, CARLA PSNR) but does not directly measure the disparity accuracy of the generated left-left and right-right views.
  • standard math Occlusion and out-of-frame masks generated from left-right consistency checks are reliable for evaluation.
    This is a widely used evaluation protocol (e.g., [46]) and is not specific to this paper. It is a reasonable assumption, though the masks inherit the inaccuracies of the underlying disparity estimates.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DMS:Diffusion-Based Multi-Baseline Stereo Generation for Improving Self-Supervised Depth Estimation." pith.science (2026). https://pith.science/paper/UOTGTXDB

@misc{pith2026250813091,
  author       = {Pith},
  title        = {Pith review of: DMS:Diffusion-Based Multi-Baseline Stereo Generation for Improving Self-Supervised Depth Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UOTGTXDB}},
  note         = {Machine review of arXiv:2508.13091}
}
read the original abstract

While supervised stereo matching and monocular depth estimation have advanced significantly with learning-based algorithms, self-supervised methods using stereo images as supervision signals have received relatively less focus and require further investigation. A primary challenge arises from ambiguity introduced during photometric reconstruction, particularly due to missing corresponding pixels in ill-posed regions of the target view, such as occlusions and out-of-frame areas. To address this and establish explicit photometric correspondences, we propose DMS, a model-agnostic approach that utilizes geometric priors from diffusion models to synthesize novel views along the epipolar direction, guided by directional prompts. Specifically, we finetune a Stable Diffusion model to simulate perspectives at key positions: left-left view shifted from the left camera, right-right view shifted from the right camera, along with an additional novel view between the left and right cameras. These synthesized views supplement occluded pixels, enabling explicit photometric reconstruction. Our proposed DMS is a cost-free, ''plug-and-play'' method that seamlessly enhances self-supervised stereo matching and monocular depth estimation, and relies solely on unlabeled stereo image pairs for both training and synthesizing. Extensive experiments demonstrate the effectiveness of our approach, with up to 35% outlier reduction and state-of-the-art performance across multiple benchmark datasets.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

98 extracted references · 42 canonical work pages

  1. [1]

    Re- imagine the negative prompt algorithm: Transform 2d diffusion into 3d, alleviate janus problem and beyond

    Mohammadreza Armandpour, Huangjie Zheng, Ali Sadeghian, Amir Sadeghian, and Mingyuan Zhou. Re- imagine the negative prompt algorithm: Transform 2d diffusion into 3d, alleviate janus problem and beyond. arXiv preprint arXiv:2304.04968, 2023. 3

    arXiv 2023

  2. [2]

    Deep 3D-Zoom Net: Unsupervised Learning of Photo-Realistic 3D-Zoom

    Juan Luis Gonzalez Bello and Munchurl Kim. Deep 3d- zoom net: Unsupervised learning of photo-realistic 3d- zoom. arXiv preprint arXiv:1909.09349, 2019. 2

    work page Pith review arXiv 1909

  3. [3]

    Deep 3d pan via local adaptive” t-shaped” convolutions with global and local adaptive dilations

    Juan Luis Gonzalez Bello and Munchurl Kim. Deep 3d pan via local adaptive” t-shaped” convolutions with global and local adaptive dilations. In International Conference on Learning Representations, 2020. 2

    2020

  4. [4]

    Zoedepth: Zero-shot trans- fer by combining relative and metric depth

    Shariq Farooq Bhat, Reiner Birkl, Diana Wofk, Peter Wonka, and Matthias M ¨uller. Zoedepth: Zero-shot trans- fer by combining relative and metric depth. arXiv preprint arXiv:2302.12288, 2023. 1

    arXiv 2023

  5. [5]

    Depth pro: Sharp monocular metric depth in less than a second

    Aleksei Bochkovskii, Ama ¨el Delaunoy, Hugo Germain, Marcel Santos, Yichao Zhou, Stephan R Richter, and Vladlen Koltun. Depth pro: Sharp monocular metric depth in less than a second. arXiv preprint arXiv:2410.02073, 2024. 1

    arXiv 2024

  6. [6]

    D. J. Butler, J. Wulff, G. B. Stanley, and M. J. Black. A naturalistic open source movie for optical flow evaluation. In Proc. Eur. Conf. Comp. Vis. , pages 611–625. Springer- Verlag, 2012. 5, 6, 7, 3

    2012

  7. [7]

    Generative novel view synthesis with 3d-aware diffusion models

    Eric R Chan, Koki Nagano, Matthew A Chan, Alexander W Bergman, Jeong Joon Park, Axel Levy, Miika Aittala, Shalini De Mello, Tero Karras, and Gordon Wetzstein. Generative novel view synthesis with 3d-aware diffusion models. In ICCV, 2023. 3

    2023

  8. [8]

    Chang and Y

    J. Chang and Y . Chen. Pyramid stereo matching network. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 5410– 5418, 2018. 1

    2018

Show all 98 references
  1. [9]

    Fantasia3d: Disentangling geometry and appearance for high-quality text-to-3d content creation

    Rui Chen, Yongwei Chen, Ningxin Jiao, and Kui Jia. Fantasia3d: Disentangling geometry and appearance for high-quality text-to-3d content creation. arXiv preprint arXiv:2303.13873, 2023. 3

    2023 arXiv

  2. [10]

    It3d: Improved text- to-3d generation with explicit view synthesis

    Yiwen Chen, Chi Zhang, Xiaofeng Yang, Zhongang Cai, Gang Yu, Lei Yang, and Guosheng Lin. It3d: Improved text- to-3d generation with explicit view synthesis. arXiv preprint arXiv:2308.11473, 2023. 3

    2023 arXiv

  3. [11]

    Hierarchical neural architecture search for deep stereo matching

    Xuelian Cheng, Yiran Zhong, Mehrtash Harandi, Yuchao Dai, Xiaojun Chang, Hongdong Li, Tom Drummond, and Zongyuan Ge. Hierarchical neural architecture search for deep stereo matching. Advances in neural information pro- cessing systems, 33:22158–22169, 2020. 2

    2020

  4. [12]

    Objaverse: A universe of annotated 3d objects

    Matt Deitke, Dustin Schwenk, Jordi Salvador, Luca Weihs, Oscar Michel, Eli VanderBilt, Ludwig Schmidt, Kiana Ehsani, Aniruddha Kembhavi, and Ali Farhadi. Objaverse: A universe of annotated 3d objects. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Rec...

    2023

  5. [13]

    Nerdi: Single-view nerf synthesis with language-guided dif- fusion as general image priors

    Congyue Deng, Chiyu Jiang, Charles R Qi, Xinchen Yan, Yin Zhou, Leonidas Guibas, Dragomir Anguelov, et al. Nerdi: Single-view nerf synthesis with language-guided dif- fusion as general image priors. In CVPR, 2023. 3

    2023

  6. [14]

    Mv-diffusion: Motion-aware video diffusion model

    Zijun Deng, Xiangteng He, Yuxin Peng, Xiongwei Zhu, and Lele Cheng. Mv-diffusion: Motion-aware video diffusion model. In Proceedings of the 31st ACM International Con- ference on Multimedia, pages 7255–7263, 2023. 3

    2023

  7. [15]

    Carla: An open urban driv- ing simulator

    Alexey Dosovitskiy, German Ros, Felipe Codevilla, Anto- nio Lopez, and Vladlen Koltun. Carla: An open urban driv- ing simulator. In Conference on robot learning, pages 1–16. PMLR, 2017. 4, 6, 1, 3

    2017

  8. [16]

    Depth map prediction from a single image using a multi-scale deep net- work

    David Eigen, Christian Puhrsch, and Rob Fergus. Depth map prediction from a single image using a multi-scale deep net- work. Advances in neural information processing systems , 27, 2014. 5, 3

    2014

  9. [17]

    Occlusion-aware self- supervised stereo matching with confidence guided raw dis- parity fusion

    Xiule Fan, Soo Jeon, and Baris Fidan. Occlusion-aware self- supervised stereo matching with confidence guided raw dis- parity fusion. In 2022 19th Conference on Robots and Vision (CRV), pages 132–139, 2022. 1, 2

    2022

  10. [18]

    Mc-stereo: Multi-peak lookup and cascade search range for stereo matching

    Miaojie Feng, Junda Cheng, Hao Jia, Longliang Liu, Gang- wei Xu, and Xin Yang. Mc-stereo: Multi-peak lookup and cascade search range for stereo matching. In 2024 Inter- national Conference on 3D Vision (3DV) , pages 344–353. IEEE, 2024. 7, 8

    2024

  11. [19]

    Are we ready for autonomous driving? the kitti vision benchmark suite

    Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the kitti vision benchmark suite. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 3354–3361. IEEE, 2012. 5, 6, 3

    2012

  12. [20]

    Vision meets robotics: The kitti dataset

    Andreas Geiger, Philip Lenz, Christoph Stiller, and Raquel Urtasun. Vision meets robotics: The kitti dataset. Interna- tional Journal of Robotics Research (IJRR), 2013. 1

    2013

  13. [21]

    Unsupervised monocular depth estimation with left- right consistency

    Cl ´ement Godard, Oisin Mac Aodha, and Gabriel J Bros- tow. Unsupervised monocular depth estimation with left- right consistency. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 270–279, 2017. 2, 5, 8

    2017

  14. [22]

    Digging into self-supervised monocular depth estimation

    Cl ´ement Godard, Oisin Mac Aodha, Michael Firman, and Gabriel J Brostow. Digging into self-supervised monocular depth estimation. In Proc. IEEE Int. Conf. Comp. Vis., pages 3828–3838, 2019. 2, 5, 8

    2019

  15. [23]

    Deep 3d pan via local adaptive” t-shaped” convolutions with global and local adaptive dilations

    Juan Luis Gonzalez Bello and Munchurl Kim. Deep 3d pan via local adaptive” t-shaped” convolutions with global and local adaptive dilations. In International Conference on Learning Representations, ICLR 2020, pages 1–20. 2020 In- ternational Conference on Learning Representatio...

    2020

  16. [24]

    Nerfdiff: Single-image view synthesis with nerf-guided dis- tillation from 3d-aware diffusion

    Jiatao Gu, Alex Trevithick, Kai-En Lin, Joshua M Susskind, Christian Theobalt, Lingjie Liu, and Ravi Ramamoorthi. Nerfdiff: Single-image view synthesis with nerf-guided dis- tillation from 3d-aware diffusion. In ICML, 2023. 3

    2023

  17. [25]

    Neural markov random field for stereo matching

    Tongfan Guan, Chen Wang, and Yun-Hui Liu. Neural markov random field for stereo matching. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5459–5469, 2024. 6

    2024

  18. [26]

    3d packing for self-supervised monocular depth estimation

    Vitor Guizilini, Rares Ambrus, Sudeep Pillai, Allan Raven- tos, and Adrien Gaidon. 3d packing for self-supervised monocular depth estimation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2485–2494, 2020. 2

    2020

  19. [27]

    X. Guo, K. Yang, W. Yang, X. Wang, and H. Li. Group-wise correlation stereo network. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 3268–3277, 2019. 1

    2019

  20. [28]

    Stereo processing by semiglobal match- ing and mutual information

    Heiko Hirschmuller. Stereo processing by semiglobal match- ing and mutual information. IEEE Trans. Pattern Anal. Mach. Intell., 30(2):328–341, 2007. 7, 8

    2007

  21. [29]

    Denoising dif- fusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. Proc. Advances in Neural Inf. Process. Syst., 33:6840–6851, 2020. 3, 5

    2020

  22. [30]

    Em- bedded real-time multi-baseline stereo

    Dominik Honegger, Torsten Sattler, and Marc Pollefeys. Em- bedded real-time multi-baseline stereo. In 2017 IEEE In- ternational Conference on Robotics and Automation (ICRA), pages 5245–5250, 2017. 2

    2017

  23. [31]

    Dreamtime: An improved optimiza- tion strategy for text-to-3d content creation

    Yukun Huang, Jianan Wang, Yukai Shi, Xianbiao Qi, Zheng- Jun Zha, and Lei Zhang. Dreamtime: An improved optimiza- tion strategy for text-to-3d content creation. arXiv preprint arXiv:2306.12422, 2023. 3

    2023 arXiv

  24. [32]

    Unsupervised monocular depth estimation with multi-baseline stereo

    Saad Imran, Muhammad Umar Karim Khan, Sikander Mukaram, and Chong-Min Kyung. Unsupervised monocular depth estimation with multi-baseline stereo. In Proc. British Machine Vis. Conf., 2020. 2

    2020

  25. [33]

    Repurpos- ing diffusion-based image generators for monocular depth estimation

    Bingxin Ke, Anton Obukhov, Shengyu Huang, Nando Met- zger, Rodrigo Caye Daudt, and Konrad Schindler. Repurpos- ing diffusion-based image generators for monocular depth estimation. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. ,

  26. [34]

    Stereonet: Guided hierarchical refinement for real-time edge-aware depth prediction

    Sameh Khamis, Sean Fanello, Christoph Rhemann, Adarsh Kowdle, Julien Valentin, and Shahram Izadi. Stereonet: Guided hierarchical refinement for real-time edge-aware depth prediction. In Proc. Eur. Conf. Comp. Vis., pages 573– 590, 2018. 7

    2018

  27. [35]

    Generative scene syn- thesis via incremental view inpainting using rgbd diffusion models

    Jiabao Lei, Jiapeng Tang, and Kui Jia. Generative scene syn- thesis via incremental view inpainting using rgbd diffusion models. In CVPR, 2022. 3

    2022

  28. [36]

    Occlusion aware stereo matching via cooperative unsupervised learning

    Ang Li and Zejian Yuan. Occlusion aware stereo matching via cooperative unsupervised learning. In Proc. Asian Conf. Comp. Vis., pages 197–213. Springer, 2018. 1, 2, 7

    2018

  29. [37]

    Era3d: High-resolution multiview diffusion using efficient row-wise attention

    Peng Li, Yuan Liu, Xiaoxiao Long, Feihu Zhang, Cheng Lin, Mengfei Li, Xingqun Qi, Shanghang Zhang, Wenhan Luo, Ping Tan, et al. Era3d: High-resolution multiview diffusion using efficient row-wise attention. arXiv preprint arXiv:2405.11616, 2024. 3

    2024 arXiv

  30. [38]

    Magic3d: High-resolution text-to-3d content creation

    Chen-Hsuan Lin et al. Magic3d: High-resolution text-to-3d content creation. arXiv preprint arXiv:2302.01335, 2023. 3

    2023 arXiv

  31. [39]

    Raft-stereo: Multilevel recurrent field transforms for stereo matching

    Lahav Lipson, Zachary Teed, and Jia Deng. Raft-stereo: Multilevel recurrent field transforms for stereo matching. In 2021 International Conference on 3D Vision (3DV) , pages 218–227. IEEE, 2021. 7, 8, 2

    2021

  32. [40]

    Geometry-aware deep network for single-image novel view synthesis

    Miaomiao Liu, Xuming He, and Mathieu Salzmann. Geometry-aware deep network for single-image novel view synthesis. In Proceedings of the IEEE Conference on Com- puter Vision and Pattern Recognition , pages 4616–4624,

  33. [41]

    Flow2stereo: Effective self-supervised learning of optical flow and stereo matching

    Pengpeng Liu, Irwin King, Michael R Lyu, and Jia Xu. Flow2stereo: Effective self-supervised learning of optical flow and stereo matching. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 6648–6657, 2020. 2, 7

    2020

  34. [42]

    Zero-1-to-3: Zero-shot one image to 3d object, 2023

    Ruoshi Liu, Rundi Wu, Basile Van Hoorick, Pavel Tok- makov, Sergey Zakharov, and Carl V ondrick. Zero-1-to-3: Zero-shot one image to 3d object, 2023. 3

    2023

  35. [43]

    Deceptive-nerf: Enhancing nerf recon- struction using pseudo-observations from diffusion models

    Xinhang Liu, Shiu-hong Kao, Jiaben Chen, Yu-Wing Tai, and Chi-Keung Tang. Deceptive-nerf: Enhancing nerf recon- struction using pseudo-observations from diffusion models. arXiv preprint arXiv:2305.15171, 2023. 3

    2023 arXiv

  36. [44]

    Syncdreamer: Gen- erating multiview-consistent images from a single-view im- age

    Yuan Liu, Cheng Lin, Zijiao Zeng, Xiaoxiao Long, Lingjie Liu, Taku Komura, and Wenping Wang. Syncdreamer: Gen- erating multiview-consistent images from a single-view im- age. arXiv preprint arXiv:2309.03453, 2023. 3

    2023 arXiv

  37. [45]

    Digging into normal incorporated stereo match- ing

    Zihua Liu, Songyan Zhang, Zhicheng Wang, and Masatoshi Okutomi. Digging into normal incorporated stereo match- ing. In Proc. ACM Int. Conf. Multimedia, pages 6050–6060,

  38. [46]

    Global occlusion-aware transformer for robust stereo matching

    Zihua Liu, Yizhou Li, and Masatoshi Okutomi. Global occlusion-aware transformer for robust stereo matching. In Proc. Winter Conf. on Appl. of Comp. Vis., pages 3535–3544,

  39. [47]

    A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation

    Nikolaus Mayer, Eddy Ilg, Philip Hausser, Philipp Fischer, Daniel Cremers, Alexey Dosovitskiy, and Thomas Brox. A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 4040–4048...

    2016

  40. [48]

    Object scene flow for autonomous vehicles

    Moritz Menze and Andreas Geiger. Object scene flow for autonomous vehicles. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 3061–3070, 2015. 5, 6, 7

    2015

  41. [49]

    Using real-time stereo vision for mobile robot navigation

    Don Murray and James J Little. Using real-time stereo vision for mobile robot navigation. Autonomous Robots , 8:161– 171, 2000. 1

    2000

  42. [50]

    Glide: Towards photorealistic image gen- eration and editing with text-guided diffusion models

    Alex Nichol et al. Glide: Towards photorealistic image gen- eration and editing with text-guided diffusion models. arXiv preprint arXiv:2112.10741, 2021. 2

    2021 arXiv

  43. [51]

    Video frame in- terpolation via adaptive separable convolution

    Simon Niklaus, Long Mai, and Feng Liu. Video frame in- terpolation via adaptive separable convolution. In Proceed- ings of the IEEE international conference on computer vi- sion, pages 261–270, 2017. 6

    2017

  44. [52]

    Video frame in- terpolation via adaptive separable convolution

    Simon Niklaus, Long Mai, and Feng Liu. Video frame in- terpolation via adaptive separable convolution. In Proceed- ings of the IEEE international conference on computer vi- sion, pages 261–270, 2017. 2

    2017

  45. [53]

    A multiple-baseline stereo

    Masatoshi Okutomi and Takeo Kanade. A multiple-baseline stereo. IEEE Transactions on pattern analysis and machine intelligence, 15(4):353–363, 1993. 2

    1993

  46. [54]

    Dreamfu- sion: Text-to-3d using 2d diffusion models

    Ben Poole, Ajay Jain, Pieter Abbeel, et al. Dreamfu- sion: Text-to-3d using 2d diffusion models. arXiv preprint arXiv:2209.14988, 2022. 3

    2022 arXiv

  47. [55]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 10684–10695, 2022. 2, 3, 5, 1

    2022

  48. [56]

    Laion-5b: An open large-scale dataset for training next generation image-text models

    Christoph Schuhmann, Romain Beaumont, Richard Vencu, Cade Gordon, Ross Wightman, Mehdi Cherti, Theo Coombes, Aarush Katta, Clayton Mullis, Mitchell Worts- man, et al. Laion-5b: An open large-scale dataset for training next generation image-text models. Advances in Neural In- f...

    2022

  49. [57]

    Ditto-nerf: Diffusion-based iterative text to omni- directional 3d model

    Hoigi Seo, Hayeon Kim, Gwanghyun Kim, and Se Young Chun. Ditto-nerf: Diffusion-based iterative text to omni- directional 3d model. arXiv preprint arXiv:2304.02827 ,

    arXiv

  50. [58]

    Let 2d diffusion model know 3d- consistency for robust text-to-3d generation

    Junyoung Seo, Wooseok Jang, Min-Seop Kwak, Jaehoon Ko, Hyeonsu Kim, Junho Kim, Jin-Hwa Kim, Jiyoung Lee, and Seungryong Kim. Let 2d diffusion model know 3d- consistency for robust text-to-3d generation. arXiv preprint arXiv:2303.07937, 2023. 3

    2023 arXiv

  51. [59]

    Cfnet: Cascade and fused cost volume for robust stereo matching

    Zhelun Shen, Yuchao Dai, and Zhibo Rao. Cfnet: Cascade and fused cost volume for robust stereo matching. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. , pages 13906–13915,

  52. [60]

    Denoising diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502, 2020. 3

    2010 arXiv

  53. [61]

    Vision-based markerless registration using stereo vi- sion and an augmented reality surgical navigation system: a pilot study

    Hideyuki Suenaga, Huy Hoang Tran, Hongen Liao, Ken Masamune, Takeyoshi Dohi, Kazuto Hoshi, and Tsuyoshi Takato. Vision-based markerless registration using stereo vi- sion and an augmented reality surgical navigation system: a pilot study. BMC Medical Imaging, 15(1):1–11, 2015. 1

    2015

  54. [62]

    Viewset diffusion:(0-) image-conditioned 3d gener- ative models from 2d data.arXiv preprint arXiv:2306.07881,

    Stanislaw Szymanowicz, Christian Rupprecht, and Andrea Vedaldi. Viewset diffusion:(0-) image-conditioned 3d gener- ative models from 2d data.arXiv preprint arXiv:2306.07881,

    arXiv

  55. [63]

    Mvdiffusion: Enabling holistic multi- view image generation with correspondence-aware diffusion

    Shitao Tang, Fuyang Zhang, Jiacheng Chen, Peng Wang, and Yasutaka Furukawa. Mvdiffusion: Enabling holistic multi- view image generation with correspondence-aware diffusion. arXiv preprint arXiv:2307.01097, 2023

    2023 arXiv

  56. [64]

    Diffusion with forward models: Solving stochastic inverse problems without direct supervision

    Ayush Tewari, Tianwei Yin, George Cazenavette, Semon Rezchikov, Joshua B Tenenbaum, Fr´edo Durand, William T Freeman, and Vincent Sitzmann. Diffusion with forward models: Solving stochastic inverse problems without direct supervision. arXiv preprint arXiv:2306.11719, 2023. 3

    2023 arXiv

  57. [65]

    Textmesh: Gen- eration of realistic 3d meshes from text prompts

    Christina Tsalicoglou, Fabian Manhardt, Alessio Tonioni, Michael Niemeyer, and Federico Tombari. Textmesh: Gen- eration of realistic 3d meshes from text prompts. arXiv preprint arXiv:2304.12439, 2023. 3

    2023 arXiv

  58. [66]

    Consistent view synthesis with pose-guided diffusion models

    Hung-Yu Tseng, Qinbo Li, Changil Kim, Suhib Alsisan, Jia- Bin Huang, and Johannes Kopf. Consistent view synthesis with pose-guided diffusion models. In CVPR, 2023. 3

    2023

  59. [67]

    Diffusers: State-of-the-art diffu- sion models

    Patrick von Platen, Suraj Patil, Anton Lozhkov, Pedro Cuenca, Nathan Lambert, Kashif Rasul, Mishig Davaadorj, Dhruv Nair, Sayak Paul, William Berman, Yiyi Xu, Steven Liu, and Thomas Wolf. Diffusers: State-of-the-art diffu- sion models. https://github.com/huggingface/ diffusers...

    2022

  60. [68]

    Score jacobian chaining: Lifting pretrained 2d diffusion models for 3d generation

    Haochen Wang, Xiaodan Du, Jiahao Li, Raymond A Yeh, and Greg Shakhnarovich. Score jacobian chaining: Lifting pretrained 2d diffusion models for 3d generation. In CVPR,

  61. [69]

    Quality prediction of asymmetrically distorted stereoscopic 3d images

    Jiheng Wang, Abdul Rehman, Kai Zeng, Shiqi Wang, and Zhou Wang. Quality prediction of asymmetrically distorted stereoscopic 3d images. IEEE Transactions on Image Pro- cessing, 24(11):3400–3414, 2015. 5, 6

    2015

  62. [70]

    Per- ceptual depth quality in distorted stereoscopic images

    Jiheng Wang, Shiqi Wang, Kede Ma, and Zhou Wang. Per- ceptual depth quality in distorted stereoscopic images. IEEE Transactions on Image Processing, 26(3):1202–1215, 2016. 5, 6

    2016

  63. [71]

    Parallax attention for unsupervised stereo correspondence learning

    Longguang Wang, Yulan Guo, Yingqian Wang, Zhengfa Liang, Zaiping Lin, Jungang Yang, and Wei An. Parallax attention for unsupervised stereo correspondence learning. IEEE Trans. Pattern Anal. Mach. Intell. , 44(4):2108–2125,

  64. [72]

    Image quality assessment: from error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli. Image quality assessment: from error visibility to structural similarity. IEEE Trans. Image Process., 13(4): 600–612, 2004. 5

    2004

  65. [73]

    Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distilla- tion

    Zhengyi Wang, Cheng Lu, Yikai Wang, Fan Bao, Chongxuan Li, Hang Su, and Jun Zhu. Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distilla- tion. arXiv preprint arXiv:2305.16213, 2023. 3

    2023 arXiv

  66. [74]

    Novel view synthesis with diffusion models

    Daniel Watson, William Chan, Ricardo Martin-Brualla, Jonathan Ho, Andrea Tagliasacchi, and Mohammad Norouzi. Novel view synthesis with diffusion models. arXiv preprint arXiv:2210.04628, 2022. 3

    2022 arXiv

  67. [75]

    Self-supervised monocular depth hints

    Jamie Watson, Michael Firman, Gabriel J Brostow, and Daniyar Turmukhambetov. Self-supervised monocular depth hints. In Proceedings of the IEEE/CVF international confer- ence on computer vision, pages 2162–2171, 2019. 2

    2019

  68. [76]

    Hd- fusion: Detailed text-to-3d generation leveraging multiple noise estimation

    Jinbo Wu, Xiaobo Gao, Xing Liu, Zhengyang Shen, Chen Zhao, Haocheng Feng, Jingtuo Liu, and Errui Ding. Hd- fusion: Detailed text-to-3d generation leveraging multiple noise estimation. arXiv preprint arXiv:2307.16183 , 2023. 3

    2023 arXiv

  69. [77]

    3d-aware image generation using 2d diffusion mod- els

    Jianfeng Xiang, Jiaolong Yang, Binbin Huang, and Xin Tong. 3d-aware image generation using 2d diffusion mod- els. arXiv preprint arXiv:2303.17905, 2023. 3

    2023 arXiv

  70. [79]

    Deep3d: Fully automatic 2d-to-3d video conversion with deep convo- lutional neural networks

    Junyuan Xie, Ross Girshick, and Ali Farhadi. Deep3d: Fully automatic 2d-to-3d video conversion with deep convo- lutional neural networks. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part IV 14 , pages 842–

    2016

  71. [80]

    Iterative geometry encoding volume for stereo matching

    Gangwei Xu, Xianqi Wang, Xiaohuan Ding, and Xin Yang. Iterative geometry encoding volume for stereo matching. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. , pages 21919– 21928, 2023. 1, 6, 7, 8, 2

    2023

  72. [81]

    Aanet: Adaptive aggregation network for efficient stereo matching

    Haofei Xu and Juyong Zhang. Aanet: Adaptive aggregation network for efficient stereo matching. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 1959–1968, 2020. 5, 1, 2

    1959

  73. [82]

    Segstereo: Exploiting semantic infor- mation for disparity estimation

    Guorun Yang, Hengshuang Zhao, Jianping Shi, Zhidong Deng, and Jiaya Jia. Segstereo: Exploiting semantic infor- mation for disparity estimation. In Proc. Eur. Conf. Comp. Vis., pages 636–651, 2018. 1, 2, 7

    2018

  74. [83]

    Drivingstereo: A large-scale dataset for stereo matching in autonomous driving scenarios

    Guorun Yang, Xiao Song, Chaoqin Huang, Zhidong Deng, Jianping Shi, and Bolei Zhou. Drivingstereo: A large-scale dataset for stereo matching in autonomous driving scenarios. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. , pages 899– 908, 2019. 1

    2019

  75. [84]

    Depth anything: Unleashing the power of large-scale unlabeled data

    Lihe Yang, Bingyi Kang, Zilong Huang, Xiaogang Xu, Jiashi Feng, and Hengshuang Zhao. Depth anything: Unleashing the power of large-scale unlabeled data. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10371–10381, 2024. 1

    2024

  76. [85]

    Depth any- thing v2

    Lihe Yang, Bingyi Kang, Zilong Huang, Zhen Zhao, Xiao- gang Xu, Jiashi Feng, and Hengshuang Zhao. Depth any- thing v2. arXiv preprint arXiv:2406.09414, 2024

    2024 arXiv

  77. [86]

    Metric3d: Towards zero-shot metric 3d prediction from a single image

    Wei Yin, Chi Zhang, Hao Chen, Zhipeng Cai, Gang Yu, Kaixuan Wang, Xiaozhi Chen, and Chunhua Shen. Metric3d: Towards zero-shot metric 3d prediction from a single image. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 9043–9053, 2023. 1

    2023

  78. [87]

    Dreamsparse: Escaping from plato’s cave with 2d frozen diffusion model given sparse views

    Paul Yoo, Jiaxian Guo, Yutaka Matsuo, and Shixiang Shane Gu. Dreamsparse: Escaping from plato’s cave with 2d frozen diffusion model given sparse views. CoRR, 2023. 3

    2023

  79. [88]

    Points-to-3d: Bridging the gap be- tween sparse points and shape-controllable text-to-3d gener- ation

    Chaohui Yu, Qiang Zhou, Jingliang Li, Zhe Zhang, Zhibin Wang, and Fan Wang. Points-to-3d: Bridging the gap be- tween sparse points and shape-controllable text-to-3d gener- ation. arXiv preprint arXiv:2307.13908, 2023. 3

    2023 arXiv

  80. [89]

    Back to basics: Unsupervised learning of optical flow via brightness constancy and motion smoothness

    Jason J Yu, Adam W Harley, and Konstantinos G Derpanis. Back to basics: Unsupervised learning of optical flow via brightness constancy and motion smoothness. In Computer Vision–ECCV 2016 Workshops: Amsterdam, The Nether- lands, October 8-10 and 15-16, 2016, Proceedings, Part I...

    2016

  81. [90]

    Yu, Fereshteh Forghani, Konstantinos G

    Jason J. Yu, Fereshteh Forghani, Konstantinos G. Derpanis, and Marcus A. Brubaker. Long-term photometric consistent novel view synthesis with diffusion models. In ICCV, 2023. 3

    2023

  82. [91]

    Zhang, V

    F. Zhang, V . Prisacariu, R. Yang, and P. H. S. Torr. Ga-net: Guided aggregation net for end-to-end stereo matching. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., 2019. 1

    2019

  83. [92]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 3836–3847, 2023. 3

    2023

  84. [93]

    Ednet: Efficient disparity estimation with cost volume combination and attention-based spatial residual

    Songyan Zhang, Zhicheng Wang, Qiang Wang, Jinshuo Zhang, Gang Wei, and Xiaowen Chu. Ednet: Efficient disparity estimation with cost volume combination and attention-based spatial residual. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn., pages 5429–5438, 2021. 1

    2021

  85. [94]

    Loss functions for image restoration with neural networks

    Hang Zhao, Orazio Gallo, Iuri Frosio, and Jan Kautz. Loss functions for image restoration with neural networks. IEEE Transactions on computational imaging , 3(1):47–57, 2016. 5

    2016

  86. [95]

    Un- supervised learning of stereo matching

    Chao Zhou, Hong Zhang, Xiaoyong Shen, and Jiaya Jia. Un- supervised learning of stereo matching. In Proc. IEEE Int. Conf. Comp. Vis., pages 1567–1575, 2017. 1, 2, 6, 7

    2017

  87. [96]

    Self- supervised monocular depth estimation with internal feature fusion

    Hang Zhou, David Greenwood, and Sarah Taylor. Self- supervised monocular depth estimation with internal feature fusion. arXiv preprint arXiv:2110.09482, 2021. 2, 8

    2021 arXiv

  88. [97]

    Self-distilled feature ag- gregation for self-supervised monocular depth estimation

    Zhengming Zhou and Qiulei Dong. Self-distilled feature ag- gregation for self-supervised monocular depth estimation. In European Conference on Computer Vision, pages 709–726. Springer, 2022. 2, 8

    2022

  89. [98]

    Sparsefusion: Dis- tilling view-conditioned diffusion for 3d reconstruction

    Zhizhuo Zhou and Shubham Tulsiani. Sparsefusion: Dis- tilling view-conditioned diffusion for 3d reconstruction. In CVPR, 2023. 3

    2023

  90. [99]

    Hifa: High-fidelity text- to-3d with advanced diffusion guidance

    Joseph Zhu and Peiye Zhuang. Hifa: High-fidelity text- to-3d with advanced diffusion guidance. arXiv preprint arXiv:2305.18766, 2023. 3

    2023 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.