Pith. sign in

REVIEW 3 major objections 6 minor 3 cited by

SharpDepth: Sharpening Metric Depth Predictions Using Diffusion Distillation

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read SharpDepth claims a diffusion-based refiner can add sharp boundaries to metric depth predictions while preserving absolute scale, using no ground-truth depth during training.

desk verdict SharpDepth has a plausible training recipe, but its own boundary-completion numbers show it losing to the trivial UniDepth-aligned Lotus baseline, so the simultaneous accuracy-and-sharpness claim doesn't hold as stated. read the letter →

arxiv 2411.18229 v1 pith:THR45XT5 submitted 2024-11-27 cs.CV

Duc-Hai Pham , Tung Do , Phong Nguyen , Binh-Son Hua , Khoi Nguyen , Rang Nguyen This is my paper · ORCID
classification cs.CV
keywords monocularmetricdepthestimationdiffusionmodelsrefinementzero-shotscoredistillationsamplingboundarysharpnessdifferencemapgatingground-truth-freetraining
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

SharpDepth claims that the two families of monocular depth estimators—discriminative models that predict accurate metric depth but smooth over fine structure, and generative diffusion models that predict sharp relative depth but lack scale—can be combined into a single zero-shot metric depth network. The paper builds a diffusion-based refiner on top of an existing metric depth model and an affine-invariant diffusion depth model, and trains it with no ground-truth depth at all. The reported results show accuracy comparable to the metric backbone while depth-boundary errors drop substantially, so thin structures like fences and poles survive in the final depth map. If the claim holds, high-quality metric depth for AR, robotics, and reconstruction can be obtained from a small, label-free training set instead of millions of annotated or synthetic images.

What carries the argument

The load-bearing mechanism is Noise-aware Gating combined with the difference map. The difference map $e$ is computed between the normalized UniDepth and Lotus predictions; the latent of the metric depth is then blended as $z'_d = \hat{e} \odot \epsilon + (1-\hat{e}) \odot z_d$, so high-disagreement regions receive more noise and become the focus of the diffusion sharpener. Two losses drive training: a modified Score Distillation Sampling (SDS) loss distills fine detail from the pretrained diffusion depth model, and a Noise-aware Reconstruction loss $L_{\text{recons}} = \| e \odot (\hat{d} - d) \|$ keeps the output anchored to the metric estimate. An exponential moving average of the training model replaces the static teacher as training progresses, so the difference map shrinks as the sharpener improves.

What would settle it

Run SharpDepth on a domain where UniDepth is known to be systematically wrong—for example, highly reflective or transparent surfaces such as those in the Booster benchmark—and compare its aligned RMSE against UniDepth's; if SharpDepth's error tracks UniDepth's error one-to-one and no shared error is ever corrected, the claim that the refiner adds independent metric accuracy is refuted.

Watch

Extended reading notes

Core claim

The central claim is that a diffusion-based depth sharpener can refine the output of a metric depth estimator so that the final prediction is simultaneously metrically accurate and visually sharp, without any ground-truth depth supervision. SharpDepth takes the metric depth map from UniDepth and the affine-invariant depth map from Lotus, normalizes both to a common range, and forms a difference map that marks where the two models disagree. The refiner receives a latent in which disagreement regions are heavily noised and agreement regions are mostly preserved, then is trained with a score distillation loss from the diffusion depth teacher and a noise-aware reconstruction loss that keeps the output close to the metric estimate. At test time the decoded output is rescaled by least-squares alignment to the original metric depth on low-difference pixels. The paper reports that this restores sharp boundaries and thin structures while retaining the absolute scale of the metric backbone.

Load-bearing premise

The load-bearing premise is that pixels where UniDepth and Lotus agree are reliable enough to anchor both the difference map and the final least-squares rescale, so any bias shared by the two models is inherited by SharpDepth.

Editorial extensions

If this is right

  • If SharpDepth is correct, metric depth estimators can be sharpened post hoc without re-training on ground-truth depth, making detail recovery cheap.
  • The training set of roughly 90,000 real images suggests that label-free refinement can reach competitive zero-shot accuracy with 100–150 times less data than discriminative metric models.
  • Downstream tasks that consume depth maps—point-cloud fusion, TSDF reconstruction, and Gaussian-splatting SLAM—should inherit sharper object boundaries from the improved depth maps, as demonstrated on TUM fr1/desk and KITTI reconstructions.
  • Because alignment is performed only on low-difference pixels, the final metric scale is tied to the metric backbone, so SharpDepth's accuracy cannot exceed the reliability of that backbone on the aligned pixels.

Reading between the lines

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

  • The method's metric accuracy is effectively inherited: the final least-squares rescale to UniDepth means SharpDepth can be no more metric than UniDepth on the aligned pixels, and a better or worse metric teacher would shift the reported accuracy accordingly.
  • The difference-map trust rule assumes that agreement between UniDepth and Lotus marks trustworthy pixels, so errors the two models share—for example, systematic biases from similar training data—would be invisible to the gating and passed through.
  • A natural extension is to treat the metric backbone as a swappable module: the same trained sharpener could be applied to any metric depth model, and the paper's qualitative Metric3Dv2 results suggest this transfer works without retraining.
  • Because the SDS loss encourages consistency with the diffusion teacher's distribution, the sharpness gain may be limited on scene contents the teacher was never trained on; feeding synthetic or out-of-domain images could test how much of the detail comes from the teacher prior versus the input image.
Share X Bluesky LinkedIn Reddit HN

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. SharpDepth proposes a diffusion-based depth sharpener that refines the metric depth predictions of a discriminative model (UniDepth) using the generative prior of an affine-invariant diffusion depth model (Lotus). The method is trained without ground-truth depth, using a difference map between the metric and affine-invariant predictions to drive a noise-aware gating mechanism, an SDS distillation loss, and a noise-aware reconstruction loss. The paper reports zero-shot depth accuracy on seven real datasets and boundary detail metrics on three synthetic and one real dataset, claiming that SharpDepth achieves both metric accuracy and sharp boundaries. The central evidence is Tables 1 and 2, with UniDepth-aligned Lotus introduced as a baseline that rescales Lotus to UniDepth using scale and shift.

Significance. If the claims were fully supported, the paper would make a useful contribution: a ground-truth-free, data-efficient refinement of metric depth that adds high-frequency detail while preserving absolute scale has clear practical value, and the proposed noise-aware gating and EMA-based teacher updating are reasonable design ideas. The training scheme on roughly 90k images is substantially cheaper than typical discriminative depth pipelines, which is a concrete strength. However, the evaluation as presented does not establish the central claim of simultaneous accuracy and sharpness: the sharpness metrics in Table 2 do not show an advantage over the simple UniDepth-aligned Lotus baseline, and the metric accuracy comparison is confounded by test-time least-squares alignment to UniDepth. The paper's contribution is therefore not yet convincingly demonstrated.

major comments (3)
  1. [Sec. 5.1, Implementation details] The final output is post-processed with least-squares alignment to the original UniDepth depth, restricted to pixels with minimal differences. This means the reported A.Rel, RMSE, and delta1 values in Tables 1, 2, and 6 are not the raw output of the learned model; part of the metric accuracy is enforced by construction. The claim that SharpDepth 'keeps the scale of UniDepth' is trivial under this protocol. The paper should report metrics without this test-time alignment, or alternatively show that the raw output is already metrically accurate and justify the alignment as an independent post-processing step. Without this, the consistent A.Rel advantage over UniDepth-aligned Lotus may be an artifact of the more refined alignment procedure rather than of the learned refinement.
  2. [Table 2, PDBE/DBE completion] The paper's own detail benchmark contradicts the sharpness half of the central claim. On the completion component of PDBE/DBE, SharpDepth is worse than UniDepth-aligned Lotus on all four reported datasets: Sintel 36.2 vs. 31.9, UnrealStereo4K 61.5 vs. 33.2, Spring 147.6 vs. 102.7, and iBims 13.1 vs. 11.0. Since UniDepth-aligned Lotus is the natural control (the generative prior rescaled to UniDepth without any learned refinement), the learned pipeline does not demonstrate improved boundary completeness; it is worse. The sentence in Sec. 5.2 that SharpDepth 'consistently surpass[es] the naive UniDepth-aligned Lotus baseline across all datasets' is only true for depth accuracy metrics, not for boundary completion. The authors should either provide a boundary metric on which SharpDepth clearly beats this baseline, or revise the claim to describe a trade-off rather than a simultaneous gain.
  3. [Sec. 4.1, difference map assumption] The gating mechanism and the test-time alignment both rely on the assumption that pixels where UniDepth and Lotus agree are reliable and need no refinement. This assumption can fail when the two models share systematic errors, for example on thin structures, transparent surfaces, or specular regions, and the paper provides no validation that low-difference pixels actually agree with ground truth. This is load-bearing for the method, because both the noise-aware gating and the reconstruction loss direct the model away from exactly those regions. The authors should add an analysis (e.g., on a small labeled validation set) showing that low-difference pixels are indeed trustworthy, or discuss the failure mode and why it is not fatal.
minor comments (6)
  1. [Sec. 5.1, Metrics] The PDBE metric is introduced but its accuracy and completion components are not defined in the main text; a brief definition or a pointer to the supplementary equation would help readers interpret Table 2.
  2. [Figure 2] The Pareto-frontier plot aggregates DBE Completion over three synthetic datasets, but the aggregation rule is not stated; please specify whether it is a mean over Sintel, UnrealStereo4K, and Spring, and whether iBims is included.
  3. [Eq. (5)] The SDS loss in Eq. (5) omits the weighting w_t that appears in Eq. (3); please state the exact weighting and noise schedule used in the implementation.
  4. [Eq. (6)] The norm in Eq. (6) is unspecified; if it is the L1 norm, this should be stated, as the type of norm affects the optimization behavior.
  5. [Sec. 5.1, Implementation details] The phrase 'the output is rescaled using least-squares alignment with the original metric depth' is ambiguous: it should be stated whether the least-squares fit uses only low-difference pixels or weights them, and how the difference map threshold is determined.
  6. [Sec. 5.1, Training data size] The claim that 90,000 images is '100-150 times smaller than the amount of data used for discriminative depth models' needs an explicit comparison; the reference models and their training set sizes should be named.

Circularity Check

2 steps flagged · score 6.0 of 10

Metric accuracy is inherited from UniDepth by construction: the output is least-squares rescaled to the input metric depth on low-difference pixels and Eq. (6) trains it to stay close to that same input; the sharpness claim is independently evaluated against GT edges, so circularity is partial.

  1. fitted input called prediction [Section 5.1, Implementation details]
    "Min-max normalization is applied to the input, and the output is rescaled using least-squares alignment with the original metric depth. The difference map is applied to ensure alignment only for pixels with minimal differences."

    The final metric depth map is post-processed by fitting a least-squares scale-and-shift to the input UniDepth depth on the trusted (low-difference) pixels. Reported metric errors (A.Rel, RMSE, δ1) therefore measure a quantity whose absolute scale is supplied by UniDepth at test time, not a scale learned or predicted independently by SharpDepth. The statement that SharpDepth 'retains' discriminative metric accuracy is thus a consequence of the alignment operation, i.e., the predicted metric component reduces to the input by construction. The sharpness component is not circular because DBE/PDBE are computed against ground-truth-derived depth edges.

  2. self definitional [Section 4.2, Noise-aware Reconstruction Loss and Eq. (6)]
    "we introduce an additional reconstruction loss that preserves the accuracy of the discriminative model by measuring the distance between our network’s output and the discriminative output in Eq. (6). ... Lrecons = ∥e ⊙ ( ˆd − d)∥"

    Here d is both the conditioning input to the sharpener and the regression target of the reconstruction loss. 'Preserving metric accuracy' is therefore defined as keeping the output close to d on the difference-weighted pixels; the model is explicitly trained so that the metric component stays anchored to UniDepth. This makes the metric half of the central claim true by definition of the training objective rather than by an independent empirical derivation. The sharpness objective (SDS) is separate and is measured against GT edges, so only the metric half is self-definitional.

full rationale

The circularity is confined to the metric-accuracy component of the central claim. SharpDepth's output is tied to UniDepth in two places: at training time, Eq. (6) minimizes the difference-weighted distance between the output and the input metric depth d, and at test time Sec. 5.1 rescales the decoded output to d with least-squares alignment on low-difference pixels. Consequently, the paper's claim of retaining discriminative accuracy is partly enforced by construction rather than independently demonstrated. The sharpness claim is not circular: DBE and PDBE are measured against depth edges derived from ground truth, not against UniDepth, and the comparison to the UniDepth-aligned Lotus baseline is a meaningful external control. I do not count the paper's own DBE/PDBE tables (Tab. 2), where SharpDepth has worse boundary completion than UniDepth-aligned Lotus on all four datasets, as circularity; that is an evidence/correctness concern about the sharpness half of the claim. The EMA self-teacher in Sec. 4.1 is self-referential but is initialized from an external Lotus teacher and the ablation (Tab. 3, settings E/G/H) shows frozen and EMA teachers behave comparably, so it is a bootstrap rather than a load-bearing equivalence. No self-citation or imported-uniqueness pattern is present. Overall, because the metric half reduces to a fit to the input while the sharpness half is independently evaluated, a partial-circularity score of 6 is appropriate.

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

The central claim depends on two pretrained teachers (UniDepth and Lotus), hand-chosen loss weights, and a test-time per-image alignment to UniDepth. No new physical entities are introduced. The main structural caveat is that metric accuracy is not independently derived; it is anchored to UniDepth by design.

free parameters (3)
  • lambda_SDS = 1.0
    Weights the score distillation loss; chosen by hand in Section 5.1.
  • lambda_recons = 0.3
    Weights the noise-aware reconstruction loss; chosen by hand in Section 5.1.
  • per-image scale and shift alignment = per image
    At test time the output is rescaled via least-squares to UniDepth on low-difference pixels; this directly sets the metric scale and is a fitted quantity, not a learned prediction.
assumptions (4)
  • domain assumption Agreement between normalized UniDepth and Lotus depths marks trustworthy regions; disagreement marks regions needing refinement.
    Invoked in Section 4.1, Noise-aware Gating.
  • domain assumption The pretrained Lotus model is a valid teacher for SDS distillation and provides a useful sharpness prior.
    Invoked in Section 4.2 and tested in ablations E-F.
  • domain assumption Training on 1% of six real datasets without ground truth is sufficient for zero-shot generalization across indoor and outdoor scenes.
    Invoked in Section 5.1.
  • domain assumption The SDS approximation dropping the U-Net Jacobian is valid for this depth distillation setting.
    Adopted from DreamFusion and SwiftBrush in Section 4.2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SharpDepth: Sharpening Metric Depth Predictions Using Diffusion Distillation." pith.science (2026). https://pith.science/paper/THR45XT5

@misc{pith2026241118229,
  author       = {Pith},
  title        = {Pith review of: SharpDepth: Sharpening Metric Depth Predictions Using Diffusion Distillation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/THR45XT5}},
  note         = {Machine review of arXiv:2411.18229}
}
read the original abstract

We propose SharpDepth, a novel approach to monocular metric depth estimation that combines the metric accuracy of discriminative depth estimation methods (e.g., Metric3D, UniDepth) with the fine-grained boundary sharpness typically achieved by generative methods (e.g., Marigold, Lotus). Traditional discriminative models trained on real-world data with sparse ground-truth depth can accurately predict metric depth but often produce over-smoothed or low-detail depth maps. Generative models, in contrast, are trained on synthetic data with dense ground truth, generating depth maps with sharp boundaries yet only providing relative depth with low accuracy. Our approach bridges these limitations by integrating metric accuracy with detailed boundary preservation, resulting in depth predictions that are both metrically precise and visually sharp. Our extensive zero-shot evaluations on standard depth estimation benchmarks confirm SharpDepth effectiveness, showing its ability to achieve both high depth accuracy and detailed representation, making it well-suited for applications requiring high-quality depth perception across diverse, real-world environments.

Figures

Figures reproduced from arXiv: 2411.18229 by the authors.

Figure 1
Figure 1. We present SharpDepth, a diffusion-based depth model for refining metric depth estimators, e.g., UniDepth [ [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The performance of SOTA depth estimation models in [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Our framework utilizes a diffusion-based estimator and a metric depth estimator to generate affine-invariant and metric depth [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (13 more)
Figure 4
Figure 4. Figure 4: The difference map between the Unidepth and Lotus [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Zero-shot qualitative results on unseen test samples of KITTI [12] and DIODE [42] dataset. Our method strikes a balance between depth accuracy and details. UniDepth lacks several details while UniDepth-aligned Lotus is less accurate. RGB Unidepth Our Unidepth point clo…
Figure 6
Figure 6. Figure 6: Un-projected point cloud from in-the-wild image. (Settings G-H), where we explore the impact of online (our approach) and offline models on difference map calcula￾tion during training. We compare two settings: (G) using the static Lotus model and (H) using the EMA (exp…
Figure 8
Figure 8. Figure 8: In-the-wild depth estimation from Internet images. Red indicates the close plane and blue means the far plane. 12 [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 9
Figure 9. Figure 9: In-the-wild depth estimation from images captured by a mobile phone. Red indicates the close plane and blue means the far plane. 13 [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]
Figure 10
Figure 10. Figure 10: Qualitative results on KITTI and NYUv2. SharpDepth* denotes our method when using depth by Metric3Dv2 as input. 14 [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]
Figure 11
Figure 11. Figure 11: Illustration of the depth boundary metrics on the Spring dataset. We show the depth maps and extracted boundaries for each prediction. Compared to UniDepth, our method extracts more edges due to better depth discontinuities. Compared to Lotus, our method can capture m…
Figure 12
Figure 12. Figure 12: Illustration of the depth boundary on the Sintel dataset. We show the depth maps and extracted boundaries for each prediction. 16 [PITH_FULL_IMAGE:figures/full_fig_p016_12.png]
Figure 13
Figure 13. Figure 13: Qualitative comparisons on different datasets (1/3). 17 [PITH_FULL_IMAGE:figures/full_fig_p017_13.png]
Figure 14
Figure 14. Figure 14: Qualitative comparisons on different datasets (2/3). 18 [PITH_FULL_IMAGE:figures/full_fig_p018_14.png]
Figure 15
Figure 15. Figure 15: Qualitative comparisons on different datasets (3/3). 19 [PITH_FULL_IMAGE:figures/full_fig_p019_15.png]
Figure 16
Figure 16. Figure 16: Rendering comparison on TUM fr1/desk sequence. For each method, we show the novel view rendering. Compared to UniDepth (leftmost column), using SharpDepth (middle column) can result in finer details of objects, such as the books in the first row and the game console i…
Figure 17
Figure 17. Figure 17: Multi-view scene reconstruction on KITTI dataset. We predict depth maps using UniDepth and SharpDepth for each frame and use TSDF-Fusion to generate the point cloud. SharpDepth’s point cloud achieves less shape distortion in vehicles. 21 [PITH_FULL_IMAGE:figures/full…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

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

  1. MetricHMSR:Metric Human Mesh and Scene Recovery from Monocular Images

    cs.CV 2025-06 conditional novelty 6.0 of 10

    From one image, MetricHMSR jointly estimates a metric human mesh, its global 3D position, and a corrected metric scene depth map.

  2. E3D-Bench: A Benchmark for End-to-End 3D Geometric Foundation Models

    cs.CV 2025-06 conditional novelty 6.0 of 10

    E3D-Bench compares 16 3D geometric foundation models on depth, reconstruction, pose, and view-synthesis tasks with a unified evaluation toolkit.

  3. Depth Anything at Any Condition

    cs.CV 2025-07 conditional novelty 5.0 of 10

    A fine-tuned Depth Anything V2 model using perturbation consistency and spatial distance constraints improves monocular depth estimation under adverse conditions without any labeled data.

Reference graph

Works this paper leans on

57 extracted references · 41 canonical work pages · cited by 3 Pith papers

  1. [1]

    Arkitscenes: A diverse real-world dataset for 3d indoor scene understanding using mobile rgb-d data

    Gilad Baruch, Zhuoyuan Chen, Afshin Dehghan, Tal Dimry, Yuri Feigin, Peter Fu, Thomas Gebauer, Brandon Joffe, Daniel Kurz, Arik Schwartz, et al. Arkitscenes: A diverse real-world dataset for 3d indoor scene understanding using mobile rgb-d data. arXiv preprint arXiv:2111.08897, 2021. 2, 5, 6, 9

    arXiv 2021

  2. [2]

    Zoedepth: Zero-shot transfer by com- bining relative and metric depth, 2023

    Shariq Farooq Bhat, Reiner Birkl, Diana Wofk, Peter Wonka, and Matthias M ¨uller. Zoedepth: Zero-shot transfer by com- bining relative and metric depth, 2023. 2, 3, 6, 7, 11

    work page 2023

  3. [3]

    nuscenes: A multi- modal dataset for autonomous driving

    Holger Caesar, Varun Bankiti, Alex H Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom. nuscenes: A multi- modal dataset for autonomous driving. In CVPR, 2020. 6, 9

    work page 2020

  4. [4]

    Oasis: A large-scale dataset for single image 3d in the wild

    Weifeng Chen, Shengyi Qian, David Fan, Noriyuki Kojima, Max Hamilton, and Jia Deng. Oasis: A large-scale dataset for single image 3d in the wild. In CVPR, 2020. 3

    work page 2020

  5. [5]

    Indoor scene understanding with geometric and semantic contexts

    Wongun Choi, Yu-Wei Chao, Caroline Pantofaru, and Silvio Savarese. Indoor scene understanding with geometric and semantic contexts. IJCV, 112, 2015. 1

    work page 2015

  6. [6]

    Scannet: Richly-annotated 3d reconstructions of indoor scenes

    Angela Dai, Angel X Chang, Manolis Savva, Maciej Hal- ber, Thomas Funkhouser, and Matthias Nießner. Scannet: Richly-annotated 3d reconstructions of indoor scenes. In Proceedings of the IEEE conference on computer vision and pattern recognition, 2017. 5, 9

    work page 2017

  7. [7]

    Towards real-time monocular depth estimation for robotics: A survey

    Xingshuai Dong, Matthew A Garratt, Sreenatha G Anavatti, and Hussein A Abbass. Towards real-time monocular depth estimation for robotics: A survey. IEEE Transactions on Intelligent Transportation Systems, 23(10), 2022. 1

    work page 2022

  8. [8]

    Omnidata: A scalable pipeline for making multi-task mid-level vision datasets from 3d scans

    Ainaz Eftekhar, Alexander Sax, Jitendra Malik, and Amir Zamir. Omnidata: A scalable pipeline for making multi-task mid-level vision datasets from 3d scans. In ICCV, 2021. 3

    work page 2021

Show all 57 references
  1. [9]

    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. NIPS, 27, 2014. 6

    2014

  2. [10]

    Deep ordinal regression net- work for monocular depth estimation

    Huan Fu, Mingming Gong, Chaohui Wang, Kayhan Bat- manghelich, and Dacheng Tao. Deep ordinal regression net- work for monocular depth estimation. In CVPR, 2018. 3

    2018

  3. [11]

    Geowiz- ard: Unleashing the diffusion priors for 3d geometry estima- tion from a single image

    Xiao Fu, Wei Yin, Mu Hu, Kaixuan Wang, Yuexin Ma, Ping Tan, Shaojie Shen, Dahua Lin, and Xiaoxiao Long. Geowiz- ard: Unleashing the diffusion priors for 3d geometry estima- tion from a single image. In ECCV. Springer, 2025. 3

    2025

  4. [12]

    Vision meets robotics: The kitti dataset

    Andreas Geiger, Philip Lenz, Christoph Stiller, and Raquel Urtasun. Vision meets robotics: The kitti dataset. IJRR, 32 (11), 2013. 6, 8, 9

    2013

  5. [13]

    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 ICCV, 2019. 1

    2019

  6. [14]

    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 CVPR, 2020

    2020

  7. [15]

    Full surround mon- odepth from multiple cameras

    Vitor Guizilini, Igor Vasiljevic, Rares Ambrus, Greg Shakhnarovich, and Adrien Gaidon. Full surround mon- odepth from multiple cameras. IEEE Robotics and Automa- tion Letters, 7(2), 2022. 1, 9

    2022

  8. [16]

    Towards zero-shot scale-aware monoc- ular depth estimation, 2023

    Vitor Guizilini, Igor Vasiljevic, Dian Chen, Rares Ambrus, and Adrien Gaidon. Towards zero-shot scale-aware monoc- ular depth estimation, 2023. 3

    2023

  9. [17]

    Lotus: Diffusion-based visual foundation model for high-quality dense prediction

    Jing He, Haodong Li, Wei Yin, Yixun Liang, Leheng Li, Kaiqiang Zhou, Hongbo Liu, Bingbing Liu, and Ying- Cong Chen. Lotus: Diffusion-based visual foundation model for high-quality dense prediction. arXiv preprint arXiv:2409.18124, 2024. 2, 3, 4, 6, 7, 9, 11

    2024 arXiv

  10. [18]

    Denoising diffu- sion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffu- sion probabilistic models. CVPR, 2020. 4

    2020

  11. [19]

    Metric3d v2: A versatile monocular geomet- ric foundation model for zero-shot metric depth and surface normal estimation

    Mu Hu, Wei Yin, Chi Zhang, Zhipeng Cai, Xiaoxiao Long, Hao Chen, Kaixuan Wang, Gang Yu, Chunhua Shen, and Shaojie Shen. Metric3d v2: A versatile monocular geomet- ric foundation model for zero-shot metric depth and surface normal estimation. arXiv preprint arXiv:2404.15506, 20...

    2024 arXiv

  12. [20]

    Brushnet: A plug-and-play image inpaint- ing model with decomposed dual-branch diffusion

    Xuan Ju, Xian Liu, Xintao Wang, Yuxuan Bian, Ying Shan, and Qiang Xu. Brushnet: A plug-and-play image inpaint- ing model with decomposed dual-branch diffusion. arXiv preprint arXiv:2403.06976, 2024. 4

    2024 arXiv

  13. [21]

    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 CVPR, 2024. 2, 3, 4, 6, 7, 9, 11

    2024

  14. [22]

    Evaluation of cnn-based single-image depth estimation methods

    Tobias Koch, Lukas Liebel, Friedrich Fraundorfer, and Marco Korner. Evaluation of cnn-based single-image depth estimation methods. In ECCV Workshops, 2018. 6, 9

    2018

  15. [23]

    From big to small: Multi-scale local planar guidance for monocular depth estimation

    Jin Han Lee, Myung-Kyu Han, Dong Wook Ko, and Il Hong Suh. From big to small: Multi-scale local planar guidance for monocular depth estimation. arXiv preprint arXiv:1907.10326, 2019. 3

    1907 arXiv

  16. [24]

    Megadepth: Learning single- view depth prediction from internet photos

    Zhengqi Li and Noah Snavely. Megadepth: Learning single- view depth prediction from internet photos. In CVPR, 2018. 3

    2018

  17. [25]

    Patchre- finer: Leveraging synthetic data for real-domain high- resolution monocular metric depth estimation

    Zhenyu Li, Shariq Farooq Bhat, and Peter Wonka. Patchre- finer: Leveraging synthetic data for real-domain high- resolution monocular metric depth estimation. In ECCV. Springer, 2024. 3, 6, 7, 11

    2024

  18. [26]

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

    Chen-Hsuan Lin, Jun Gao, Luming Tang, Towaki Takikawa, Xiaohui Zeng, Xun Huang, Karsten Kreis, Sanja Fidler, Ming-Yu Liu, and Tsung-Yi Lin. Magic3d: High-resolution text-to-3d content creation. In CVPR, 2023. 3

    2023

  19. [27]

    Repaint: Inpainting using denoising diffusion probabilistic models

    Andreas Lugmayr, Martin Danelljan, Andres Romero, Fisher Yu, Radu Timofte, and Luc Gool. Repaint: Inpainting using denoising diffusion probabilistic models. In CVPR, 2022. 4

    2022

  20. [28]

    Hidenobu Matsuki, Riku Murai, Paul H. J. Kelly, and An- drew J. Davison. Gaussian Splatting SLAM. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, 2024. 10

    2024

  21. [29]

    Spring: A high-resolution high- detail dataset and benchmark for scene flow, optical flow and stereo

    Lukas Mehl, Jenny Schmalfuss, Azin Jahedi, Yaroslava Nali- vayko, and Andr ´es Bruhn. Spring: A high-resolution high- detail dataset and benchmark for scene flow, optical flow and stereo. In CVPR, 2023. 2, 6, 9

    2023

  22. [30]

    Swiftbrush: One-step text-to-image diffusion model with variational score distilla- tion

    Thuan Hoang Nguyen and Anh Tran. Swiftbrush: One-step text-to-image diffusion model with variational score distilla- tion. In CVPR, 2024. 4 22

    2024

  23. [31]

    UniDepth: Universal monocular metric depth estimation

    Luigi Piccinelli, Yung-Hsu Yang, Christos Sakaridis, Mattia Segu, Siyuan Li, Luc Van Gool, and Fisher Yu. UniDepth: Universal monocular metric depth estimation. In CVPR,

  24. [32]

    Barron, and Ben Milden- hall

    Ben Poole, Ajay Jain, Jonathan T. Barron, and Ben Milden- hall. Dreamfusion: Text-to-3d using 2d diffusion. arXiv,

  25. [33]

    Booster: a benchmark for depth from images of specular and transparent surfaces

    Pierluigi Zama Ramirez, Alex Costanzino, Fabio Tosi, Mat- teo Poggi, Samuele Salti, Stefano Mattoccia, and Luigi Di Stefano. Booster: a benchmark for depth from images of specular and transparent surfaces. PAMI, 2023. 6, 9

    2023

  26. [34]

    Vi- sion transformers for dense prediction

    Ren ´e Ranftl, Alexey Bochkovskiy, and Vladlen Koltun. Vi- sion transformers for dense prediction. In ICCV, 2021. 3

    2021

  27. [35]

    Towards robust monocular depth estimation: Mixing datasets for zero-shot cross-dataset transfer

    Ren ´e Ranftl, Katrin Lasinger, David Hafner, Konrad Schindler, and Vladlen Koltun. Towards robust monocular depth estimation: Mixing datasets for zero-shot cross-dataset transfer. PAMI, 44(3), 2022. 3

    2022

  28. [36]

    High-resolution image syn- thesis with latent diffusion models, 2021

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj¨orn Ommer. High-resolution image syn- thesis with latent diffusion models, 2021. 3, 4

    2021

  29. [37]

    A multi-view stereo benchmark with high- resolution images and multi-camera videos

    Thomas Schops, Johannes L Schonberger, Silvano Galliani, Torsten Sattler, Konrad Schindler, Marc Pollefeys, and An- dreas Geiger. A multi-view stereo benchmark with high- resolution images and multi-camera videos. In CVPR, 2017. 6

    2017

  30. [38]

    Indoor segmentation and support inference from rgbd images

    Nathan Silberman, Derek Hoiem, Pushmeet Kohli, and Rob Fergus. Indoor segmentation and support inference from rgbd images. In ECCV. Springer, 2012. 6, 9

    2012

  31. [39]

    A benchmark for the evalua- tion of rgb-d slam systems

    J ¨urgen Sturm, Nikolas Engelhard, Felix Endres, Wolfram Burgard, and Daniel Cremers. A benchmark for the evalua- tion of rgb-d slam systems. In 2012 IEEE/RSJ international conference on intelligent robots and systems, pages 573–580. IEEE, 2012. 9, 10

    2012

  32. [40]

    Scalability in perception for autonomous driving: Waymo open dataset

    Pei Sun, Henrik Kretzschmar, Xerxes Dotiwalla, Aurelien Chouard, Vijaysai Patnaik, Paul Tsui, James Guo, Yin Zhou, Yuning Chai, Benjamin Caine, et al. Scalability in perception for autonomous driving: Waymo open dataset. In CVPR,

  33. [41]

    Smd-nets: Stereo mixture density networks

    Fabio Tosi, Yiyi Liao, Carolin Schmitt, and Andreas Geiger. Smd-nets: Stereo mixture density networks. In CVPR, 2021. 2, 6, 9

    2021

  34. [42]

    Diode: A dense indoor and outdoor depth dataset

    Igor Vasiljevic, Nick Kolkin, Shanyi Zhang, Ruotian Luo, Haochen Wang, Falcon Z Dai, Andrea F Daniele, Moham- madreza Mostajabi, Steven Basart, Matthew R Walter, et al. Diode: A dense indoor and outdoor depth dataset. arXiv preprint arXiv:1908.00463, 2019. 6, 8, 9

    1908 arXiv

  35. [43]

    Yeh, and Greg Shakhnarovich

    Haochen Wang, Xiaodan Du, Jiahao Li, Raymond A. Yeh, and Greg Shakhnarovich. Score jacobian chaining: Lift- ing pretrained 2d diffusion models for 3d generation. arXiv preprint arXiv:2212.00774, 2022. 3

    2022 arXiv

  36. [44]

    Can scale-consistent monocu- lar depth be learned in a self-supervised scale-invariant man- ner? In ICCV, 2021

    Lijun Wang, Yifan Wang, Linzhao Wang, Yunlong Zhan, Ying Wang, and Huchuan Lu. Can scale-consistent monocu- lar depth be learned in a self-supervised scale-invariant man- ner? In ICCV, 2021. 1

    2021

  37. [45]

    Self-supervised monocular depth hints

    Jamie Watson, Michael Firman, Gabriel J Brostow, and Daniyar Turmukhambetov. Self-supervised monocular depth hints. In ICCV, 2019. 1

    2019

  38. [46]

    Argoverse 2: Next generation datasets for self-driving perception and forecasting

    Benjamin Wilson, William Qi, Tanmay Agarwal, John Lambert, Jagjeet Singh, Siddhesh Khandelwal, Bowen Pan, Ratnesh Kumar, Andrew Hartnett, Jhony Kaesemodel Pontes, et al. Argoverse 2: Next generation datasets for self-driving perception and forecasting. arXiv preprint arXiv:230...

    2023 arXiv

  39. [47]

    Lessons and insights from creating a syn- thetic optical flow benchmark

    Jonas Wulff, Daniel J Butler, Garrett B Stanley, and Michael J Black. Lessons and insights from creating a syn- thetic optical flow benchmark. In ECCV. Springer, 2012. 2, 6, 9

    2012

  40. [48]

    Pandaset: Advanced sensor suite dataset for autonomous driving

    Pengchuan Xiao, Zhenlei Shao, Steven Hao, Zishuo Zhang, Xiaolin Chai, Judy Jiao, Zesong Li, Jian Wu, Kai Sun, Kun Jiang, et al. Pandaset: Advanced sensor suite dataset for autonomous driving. In 2021 IEEE International Intelligent Transportation Systems Conference (ITSC). IEEE...

    2021

  41. [49]

    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 CVPR, 2024. 2, 3

    2024

  42. [50]

    Virtual normal: En- forcing geometric constraints for accurate and robust depth prediction

    Wei Yin, Yifan Liu, and Chunhua Shen. Virtual normal: En- forcing geometric constraints for accurate and robust depth prediction. PAMI, 2021. 3

    2021

  43. [51]

    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 ICCV, pages 9043–9053, 2023. 3, 5

    2023

  44. [52]

    Real-time monocular depth estima- tion with sparse supervision on mobile

    Mehmet Kerim Yucel, Valia Dimaridou, Anastasios Drosou, and Albert Saa-Garriga. Real-time monocular depth estima- tion with sparse supervision on mobile. In CVPR, 2021. 1

    2021

  45. [53]

    Taskonomy: Disentangling task transfer learning

    Amir R Zamir, Alexander Sax, William Shen, Leonidas J Guibas, Jitendra Malik, and Silvio Savarese. Taskonomy: Disentangling task transfer learning. In CVPR, 2018. 5, 9

    2018

  46. [54]

    3dmatch: Learning local geometric descriptors from rgb-d reconstruc- tions

    Andy Zeng, Shuran Song, Matthias Nießner, Matthew Fisher, Jianxiong Xiao, and Thomas Funkhouser. 3dmatch: Learning local geometric descriptors from rgb-d reconstruc- tions. In CVPR, 2017. 9, 10

    2017

  47. [55]

    Betterdepth: Plug-and-play diffu- sion refiner for zero-shot monocular depth estimation

    Xiang Zhang, Bingxin Ke, Hayko Riemenschneider, Nando Metzger, Anton Obukhov, Markus Gross, Konrad Schindler, and Christopher Schroers. Betterdepth: Plug-and-play diffu- sion refiner for zero-shot monocular depth estimation. arXiv preprint arXiv:2407.17952, 2024. 3, 6, 7

    2024 arXiv

  48. [56]

    Tryondiffusion: A tale of two un- ets

    Luyang Zhu, Dawei Yang, Tyler Zhu, Fitsum Reda, William Chan, Chitwan Saharia, Mohammad Norouzi, and Ira Kemelmacher-Shlizerman. Tryondiffusion: A tale of two un- ets. In CVPR, 2023. 4 23

    2023

  49. [2024]

    1, 2, 3, 4, 5, 6, 7, 10, 11

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.