Pith. sign in

REVIEW 3 major objections 5 minor 42 references

Adversarial View-Consistent Learning for Monocular Depth Estimation

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Monocular depth networks trained with an adversarial pose generator that warps predictions into hard views produce view-consistent depth maps and outperform their single-view baselines on NYU Depth v2.

desk verdict A plausible view-consistency training trick with a geometrically sound core, undermined by inconsistent tables that need resolving before the empirical claims can be trusted. read the letter →

arxiv 1908.01301 v1 pith:RB3PG4Z4 submitted 2019-08-04 cs.CV cs.LGeess.IV

Yixuan Liu , Yuwang Wang , Shengjin Wang This is my paper · ORCID
classification cs.CVcs.LGeess.IV
keywords monoculardepthestimationview-consistentdifferentiablemapwarpingadversarialposegenerationnovelviewsynthesismulti-viewstereoNYUv2hardexamplemining
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

Monocular depth estimation is usually trained with a per-pixel loss in the original camera view, so two predictions with the same loss can still look very different from another viewpoint. This paper argues that a good depth map should be view-consistent: small error in the original view and also small error after the scene is rendered from other directions. To exploit that, the authors add a differentiable warping branch that projects the predicted depth and the ground-truth depth into a novel view, plus an adversarial pose generator that deliberately chooses the hardest viewpoint. Trained with the resulting warped loss, depth networks predict sharper edges and lower error across views, and the improvement stacks on top of existing losses such as DORN without changing inference. The same supervision lets a network reach a given accuracy with a tenth of the training data.

What carries the argument

The load-bearing mechanism is the differentiable depth-map warping operation $W(D_s, p_{s→t})$, which projects each source pixel into a target camera frame using the camera intrinsics $K$ and a 6-DoF pose, then fills each target pixel with the minimum depth among all source points that land on it, treating empty pixels as ignored. This z-buffer rule keeps the operation differentiable and geometrically plausible when surfaces occlude one another. The second component is the adversarial pose generator $G$, a network branch that takes the intermediate feature map and emits a pose vector $p$, clamped by a scaled sigmoid; it is trained with the inverted gradient of $L_{warp}$ plus a quadratic penalty $\lambda(p⊙p)$, so it learns to choose hard views instead of easy ones. During training the depth network receives gradients from both $L_{dep}$ and $L_{warp}$, while the pose generator receives only the adversarial gradient; at inference both the pose generator and the warping branch are discarded.

What would settle it

On NYU Depth v2, take a trained AVCL model, warp its predicted depth and the ground-truth depth with a fixed pose, and for every pixel that contributes to $L_{warp}$ check whether the 3D point that produced the predicted value is the same as the 3D point from the ground-truth depth. If a large share of the high-loss pixels are occlusion or correspondence mismatches, where the predicted depth lands on a different surface than the ground-truth depth at that pixel, then the loss is not enforcing view-consistent geometry as claimed.

Watch

Extended reading notes

Core claim

The central claim is that forcing consistency across multiple views is a stronger training signal than the single-view regression loss, and that the views should be chosen adversarially rather than fixed or random. The paper demonstrates this by augmenting existing monocular depth networks with an AVCL branch: the predicted depth map is warped to a target view using a differentiable projection and z-buffering operation $W$, the ground-truth depth map is warped with the same pose, and an $L_1$ loss $L_{warp}$ compares the two warped depth maps. A pose generator $G$ is trained against the depth network by inverting its gradient, so it learns to output a 6-DoF pose that makes the warped prediction disagree most with the warped ground truth. On NYU Depth v2, this raises DORN's $\delta_1$ accuracy from 0.828 to 0.836 with the full training set and improves every tested loss function on the smaller 1.2K training set.

Load-bearing premise

The argument assumes that when both the predicted depth and the ground-truth depth are warped with the same pose, the pixel at a given location in the warped prediction and the pixel at the same location in the warped ground truth correspond to the same physical surface point, so the $L_{warp}$ loss measures a real geometric disagreement rather than comparing unrelated 3D points.

Editorial extensions

If this is right

  • Any existing monocular depth network can be augmented with AVCL during training while keeping its inference-time architecture unchanged, so the gain is independent of the network's backbone or loss function.
  • Training on one tenth of NYU Depth v2 with AVCL reaches roughly the same accuracy as training on the full dataset without it, indicating that the view-consistency loss acts as data-efficient augmentation.
  • The adversarial pose selection matters: fixed poses can hurt performance depending on the pose, and random poses help less than adversarially chosen poses, so the framework transfers the burden of choosing supervision views to the learned pose generator.
  • Because the warped branch supervises the depth map as a 3D surface rather than as a pixel array, the predicted depth becomes more consistent across rendering directions, which is exactly what downstream applications such as view synthesis and 3D reconstruction need.

Reading between the lines

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

  • The same adversarial warping idea could apply to other dense prediction tasks with geometric meaning, such as surface-normal estimation or semantic 3D voxel grids, where consistency across views is a natural correctness criterion.
  • The paper does not explicitly verify that the warped predicted depth and the warped ground truth refer to the same physical surface point at each pixel; a pixel-level correspondence test would show whether the $L_{warp}$ signal is purely geometric or partly a texture-like averaging effect.
  • A testable extension is to replace the single adversarial pose with multiple poses or a sequence of warps and to enforce cycle consistency by warping back and forth, which would reveal whether hard-view mining or sheer view diversity drives the gain.
  • Because DORN's ordinal output is approximated with sigmoid sums to make it warpable, the same soft-thresholding trick could extend AVCL to other discrete or ordinal regression networks.
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 / 5 minor

Summary. The paper proposes an Adversarial View-Consistent Learning (AVCL) framework for monocular depth estimation. The core idea is to warp the predicted depth map into novel views using a proposed differentiable warping operation, compare the warped prediction against a warped ground-truth depth map with an L1 loss, and train an adversarial pose generator to select hard views that maximize this comparison loss. The framework is designed as a plug-in module that can be combined with existing depth estimation losses (L1, berHu, DORN) and networks (ResNet-based). Experiments on NYU Depth v2 are conducted under two training-data settings (full 120K images and a 1/10 down-sampled subset) and show consistent improvements: for example, DORN+AVCL reaches 0.836 δ1 accuracy versus 0.828 for DORN alone when trained on the full dataset, and the method is claimed to reduce the amount of training data needed to reach a given accuracy.

Significance. If the reported results are reliable, the paper makes a useful empirical contribution: a simple, network-agnostic regularization that improves multiple monocular depth baselines, with a tangible data-efficiency gain on a down-sampled training set. The differentiable warping operation and the adversarial hard-view mining idea are interesting building blocks for future view-consistency work. The paper's strengths are the consistent gains across three loss functions and the clear qualitative illustration of view-consistency. However, the paper does not release code or trained models, does not report error bars or multiple runs, and contains internal numerical inconsistencies that currently undermine the credibility of the ablation study.

major comments (3)
  1. [§4.2, Tables 1 and 4] The ablation table is internally inconsistent with the main comparison table. Table 4 is described as using L1 loss and the down10 training set, but its 'Baseline' entry reports δ1 = 0.817, whereas Table 1 lists the same setting (L1 Loss, no AVCL, down10) as 0.800. Moreover, the 'Adversarial pose' row of Table 4 reports 0.836, which is the same nominal configuration as 'L1 Loss ✓' in Table 1 (0.810). These numbers should coincide unless the two tables use different networks, data splits, or training protocols; the paper does not state any such difference. This discrepancy makes the ablation's conclusions about fixed/random/adversarial poses difficult to interpret and needs to be resolved before the paper can be accepted.
  2. [§3.1, Eqs. (2)-(3)] The paper's claim that Eq. (3) enforces 'view-consistency' needs a more careful geometric discussion. For a fixed target pixel (i,j), both D'_pred and D'_gt are computed as the minimum depth along the same target-view ray, but that ray hits different scene surfaces depending on the predicted depth. The L1 loss therefore is a valid novel-view depth rendering loss, not a comparison of unrelated scene points; the skeptic's correspondence-mismatch concern does not invalidate the training signal. However, the min-selection in Eq. (2) means that only the point with the smallest z contributes to each target pixel, so depth errors on surfaces that are occluded in the rendered view are never penalized. The paper does not discuss this occlusion/error-hiding behavior or report the fraction of target pixels that are ignored or correspond to different source points. This limitation is directly relevant to the adversarial pose generator, which may exploit occlusion mismatches to inflate L_warp rather than mine geometrically 'hard' views. The authors should add an analysis or at least a discussion of this behavior.
  3. [§4.2, Table 4] The superiority of the adversarial pose module over random poses rests on a single point estimate: 0.836 versus 0.825 δ1. Given that the fixed-pose ablations show substantial sensitivity (0.802–0.822), a single run is insufficient to support the claim that adversarial mining is necessary or consistently better. The paper should report multiple independent runs with means and standard deviations, or provide a paired comparison over several seeds, for the key ablation rows. Without this, the gain from adversarial pose generation is not statistically grounded.
minor comments (5)
  1. [Throughout] There are numerous typos and inconsistent notations: 'Tabel' for 'Table', 'A VCL / A VCL' spacing, 'fuctions' for 'functions', 'structral' for 'structural', 'avaible' for 'available', 'contract' for 'contrast', and inconsistent use of 'L1 Loss' versus 'L1 loss'.
  2. [§4.2, Table 4 caption] The caption 'Comparisons with existing monocular depth estimation methods on NYU Depth v2 dataset' is inaccurate; the table shows ablations of the proposed method and should be relabeled accordingly.
  3. [§3.1, Eq. (2)] The paper claims the warping operation is differentiable because it uses 'simple value assignment', but the floor operation is non-differentiable at integer boundaries and the min-selection is piecewise constant. The authors should state precisely how gradients are computed through the coordinate discretization, for example by defining an indicator-based straight-through estimator or by noting that gradients flow only to the selected source pixel.
  4. [§4.1, Implementation Details] The hyperparameter λ is set to (1,1,1,1,1,1) in all experiments, but the paper does not explain whether this vector multiplies the pose vector elementwise or its squared norm; Eq. (4) writes λ·(p⊙p), which is ambiguous if λ is a vector. Clarify the exact regularization form.
  5. [§4.2, DORN loss adaptation] For the DORN loss, the warped branch supervises a continuous approximation of the thresholded output, while the main branch uses the ordinal regression loss on the thresholded output. The authors should discuss whether the benefit could partly come from a form of auxiliary smoothed-target regression rather than from view-consistency alone.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: AVCL is an empirical training objective evaluated on an external benchmark; the adversarial pose generator is a min-max mechanism, not a fitted input renamed as a prediction.

full rationale

The paper's chain is empirical rather than derivational: Ldep and Lwarp in Eq. (3) are training losses, and the method's success is measured on NYU Depth v2 with the standard delta, rel, log10, rms, and rms(log) metrics. No parameter is fitted to the evaluation set, and no prediction is defined as the value of a fitted constant. The adversarial pose generator is trained to maximize Lwarp while the depth generator minimizes it; this min-max loop is the proposed mechanism, not a hidden reuse of the output as input. The closest self-referential element is that 'hard views' are those with the largest current warping loss, but that is a training-signal property, not a derivation of the central claim from itself. The concern that D'_pred and D'_gt in Eq. (3) may select different scene points when predicted and ground-truth depths disagree is a correctness and robustness issue about the geometric validity of the supervision, not a circularity: even if the loss is noisy, the benchmark comparison remains external and falsifiable. No load-bearing self-citations or imported uniqueness theorems appear. Thus no circular step meeting the evidence bar is present.

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

The method's central training loss relies on standard camera projection, reliable ground truth depth, the validity of the min-depth warping proxy, and the stability of the adversarial pose optimizer. The paper does not provide code or a sensitivity analysis for the pose regularization, so several of these assumptions are unverified.

free parameters (3)
  • λ pose regularization = (1,1,1,1,1,1)
    In Eq. (4), λ weights the squared pose penalty. Chosen without reported tuning or sensitivity analysis; it directly controls the range of adversarial poses, so it affects the hard-view mining behavior.
  • Pose sigmoid scale and shift = unspecified
    Section 3.2 says the pose output is restricted by a pre-defined scaled and shifted sigmoid, but the scale and shift are never given; these bound the achievable rotations and translations.
  • DORN soft-threshold width = 100 and 0.5
    For DORN loss, the paper replaces thresholding by sum_i sigmoid(100(x-0.5)); these constants are chosen manually and not ablated.
assumptions (4)
  • domain assumption The pinhole camera model with known intrinsics K (Eq. 1) correctly describes image formation for NYU Depth v2.
    The warping operation projects 2D pixels to 3D and back using K; if intrinsics vary per image, the fixed K assumption would be wrong. NYU provides aligned RGB-D with known intrinsics for downsampled images.
  • domain assumption Ground truth depth maps are accurate enough to serve as warping targets, and invalid/no-depth pixels can be safely ignored.
    NYU v2 depth is captured by Kinect, which has missing and edge-invalid regions; the loss ignores pixels that receive no depth, which may bias training if invalid regions correlate with errors.
  • ad hoc to paper The min-depth, floor-based warping in Eq. (2) approximates true depth re-rendering and provides useful gradients despite the non-differentiable floor and argmin selection.
    The paper asserts differentiability but does not analyze the straight-through gradient or the effect of multiple source pixels per target pixel on the loss.
  • ad hoc to paper The adversarial pose generator, trained with inverted gradients, will converge to hard but non-degenerate views rather than poses that escape the loss by producing mostly ignored pixels.
    No proof or analysis; λ and sigmoid bounds are the only safeguards, and the paper does not show the distribution of generated poses.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Adversarial View-Consistent Learning for Monocular Depth Estimation." pith.science (2026). https://pith.science/paper/RB3PG4Z4

@misc{pith2026190801301,
  author       = {Pith},
  title        = {Pith review of: Adversarial View-Consistent Learning for Monocular Depth Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RB3PG4Z4}},
  note         = {Machine review of arXiv:1908.01301}
}
read the original abstract

This paper addresses the problem of Monocular Depth Estimation (MDE). Existing approaches on MDE usually model it as a pixel-level regression problem, ignoring the underlying geometry property. We empirically find this may result in sub-optimal solution: while the predicted depth map presents small loss value in one specific view, it may exhibit large loss if viewed in different directions. In this paper, inspired by multi-view stereo (MVS), we propose an Adversarial View-Consistent Learning (AVCL) framework to force the estimated depth map to be all reasonable viewed from multiple views. To this end, we first design a differentiable depth map warping operation, which is end-to-end trainable, and then propose a pose generator to generate novel views for a given image in an adversarial manner. Collaborating with the differentiable depth map warping operation, the pose generator encourages the depth estimation network to learn from hard views, hence produce view-consistent depth maps . We evaluate our method on NYU Depth V2 dataset and the experimental results show promising performance gain upon state-of-the-art MDE approaches.

Figures

Figures reproduced from arXiv: 1908.01301 by the authors.

Figure 1
Figure 1. An illustration on view-consistent depth map (prediction B) and view-inconsistent [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. An overview of the proposed Adversarial View-Consistent Learning framework. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Gradient flow during training in AVCL. where Ds(ps) is the depth value of the pixel, and K denotes the camera intrinsics matrix. Rs→t ∈ R 3×3 and Ts→t ∈ R 3 are the rotation matrix and translation vector from the source view to the target view, respectively, calculated from ps→t . The whole process can be bro￾ken down into three procedures. First, the source pixel coordinates are mapped into world coordinates, i.e.,… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Qualitative comparison between with and without AVCL in different views. δ1 V1 V2 V3 V4 w/ AVCL 0.825 0.836 0.816 0.826 wo AVCL 0.818 0.827 0.808 0.820 [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 33 canonical work pages

  1. [1]

    Large scale gan training for high fidelity natural image synthesis

    Andrew Brock, Jeff Donahue, and Karen Simonyan. Large scale gan training for high fidelity natural image synthesis. arXiv preprint arXiv:1809.11096, 2018

    arXiv 2018

  2. [2]

    Depth from a single image by harmonizing overcomplete local network predictions

    Ayan Chakrabarti, Jingyu Shao, and Greg Shakhnarovich. Depth from a single image by harmonizing overcomplete local network predictions. In NIPS. 2016

    work page 2016

  3. [3]

    Predicting depth, surface normals and semantic labels with a common multi-scale convolutional architecture

    David Eigen and Rob Fergus. Predicting depth, surface normals and semantic labels with a common multi-scale convolutional architecture. In CVPR, 2015

    work page 2015

  4. [4]

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

    David Eigen, Christian Puhrsch, and Rob Fergus. Depth map prediction from a single image using a multi-scale deep network. In Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, and K. Q. Weinberger, editors,NIPS. 2014

    work page 2014

  5. [5]

    Deep ordinal regression network for monocular depth estimation

    Huan Fu, Mingming Gong, Chaohui Wang, Kayhan Batmanghelich, and Dacheng Tao. Deep ordinal regression network for monocular depth estimation. In CVPR, 2018

    2018

  6. [6]

    Multi-view stereo: A tutorial

    Yasutaka Furukawa, Carlos Hernández, et al. Multi-view stereo: A tutorial. F ounda- tions and Trends R© in Computer Graphics and Vision, 9(1-2):1–148, 2015

    work page 2015

  7. [7]

    Massively parallel multi- view stereopsis by surface normal diffusion

    Silvano Galliani, Katrin Lasinger, and Konrad Schindler. Massively parallel multi- view stereopsis by surface normal diffusion. In Proceedings of the IEEE International Conference on Computer Vision, pages 873–881, 2015

    2015

  8. [8]

    Generative adversarial nets

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In NIPS, 2014

    2014

Show all 42 references
  1. [9]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In CVPR, 2016

    2016

  2. [10]

    Recovering surface layout from an image

    Derek Hoiem, Alexei A Efros, and Martial Hebert. Recovering surface layout from an image. International Journal of Computer Vision, 75(1):151–172, 2007

    2007

  3. [11]

    Beyond face rotation: Global and local perception gan for photorealistic and identity preserving frontal view synthesis

    Rui Huang, Shu Zhang, Tianyu Li, and Ran He. Beyond face rotation: Global and local perception gan for photorealistic and identity preserving frontal view synthesis. In ICCV, 2017

    2017

  4. [12]

    Spatial transformer net- works

    Max Jaderberg, Karen Simonyan, Andrew Zisserman, et al. Spatial transformer net- works. In NIPS, 2015

    2015

  5. [13]

    Surfacenet: An end-to-end 3d neural network for multiview stereopsis

    Mengqi Ji, Juergen Gall, Haitian Zheng, Yebin Liu, and Lu Fang. Surfacenet: An end-to-end 3d neural network for multiview stereopsis. In Proceedings of the IEEE International Conference on Computer Vision, pages 2307–2315, 2017

    2017

  6. [14]

    Karsch, C

    K. Karsch, C. Liu, and S. B. Kang. Depth transfer: Depth extraction from video us- ing non-parametric sampling. IEEE Transactions on Pattern Analysis and Machine Intelligence, 36(11):2144–2158, 2014

    2014

  7. [15]

    What uncertainties do we need in bayesian deep learning for computer vision? In nips

    Alex Kendall and Yarin Gal. What uncertainties do we need in bayesian deep learning for computer vision? In nips. 2017. Y .LIU, Y .W ANG, S.W ANG: A VCL 11

    2017

  8. [16]

    A theory of shape by space carving

    Kiriakos N Kutulakos and Steven M Seitz. A theory of shape by space carving. Inter- national journal of computer vision , 38(3):199–218, 2000

    2000

  9. [17]

    Pulling things out of perspective

    Lubor Ladicky, Jianbo Shi, and Marc Pollefeys. Pulling things out of perspective. In CVPR, 2014

    2014

  10. [18]

    Deeper depth prediction with fully convolutional residual networks

    Iro Laina, Christian Rupprecht, Vasileios Belagiannis, Federico Tombari, and Nassir Navab. Deeper depth prediction with fully convolutional residual networks. In 2016 F ourth international conference on 3D vision (3DV), 2016

    2016

  11. [19]

    Depth and surface normal estimation from monocular images using regression on deep features and hierarchical crfs

    Bo Li, Chunhua Shen, Yuchao Dai, Anton van den Hengel, and Mingyi He. Depth and surface normal estimation from monocular images using regression on deep features and hierarchical crfs. In CVPR, 2015

    2015

  12. [21]

    A two-streamed network for estimating fine- scaled depth maps from single rgb images

    Jun Li, Reinhard Klein, and Angela Yao. A two-streamed network for estimating fine- scaled depth maps from single rgb images. In ICCV, 2017

    2017

  13. [22]

    B. Liu, S. Gould, and D. Koller. Single image depth estimation from predicted semantic labels. In cvpr, 2010

    2010

  14. [23]

    Ce Liu, Jenny Yuen, Antonio Torralba, Josef Sivic, and William T. Freeman. Sift flow: Dense correspondence across different scenes. In David Forsyth, Philip Torr, and Andrew Zisserman, editors, ECCV, 2008

    2008

  15. [24]

    F. Liu, C. Shen, G. Lin, and I. Reid. Learning depth from single monocular images using deep convolutional neural fields. IEEE Transactions on Pattern Analysis and Machine Intelligence, 38(10):2024–2039, 2016

    2024

  16. [26]

    Learning depth from single monocular images using deep convolutional neural fields.IEEE transactions on pattern analysis and machine intelligence, 38(10):2024–2039, 2016

    Fayao Liu, Chunhua Shen, Guosheng Lin, and Ian Reid. Learning depth from single monocular images using deep convolutional neural fields.IEEE transactions on pattern analysis and machine intelligence, 38(10):2024–2039, 2016

    2024

  17. [27]

    Discrete-continuous depth esti- mation from a single image

    Miaomiao Liu, Mathieu Salzmann, and Xuming He. Discrete-continuous depth esti- mation from a single image. In CVPR, 2014

    2014

  18. [28]

    Deep metric learn- ing with bier: Boosting independent embeddings robustly

    Michael Opitz, Georg Waltner, Horst Possegger, and Horst Bischof. Deep metric learn- ing with bier: Boosting independent embeddings robustly. IEEE transactions on pat- tern analysis and machine intelligence , 2018

    2018

  19. [29]

    Unsupervised representation learning with deep convolutional generative adversarial networks

    Alec Radford, Luke Metz, and Soumith Chintala. Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv preprint arXiv:1511.06434, 2015

    2015 arXiv

  20. [30]

    Monocular depth estimation using neural regres- sion forest

    Anirban Roy and Sinisa Todorovic. Monocular depth estimation using neural regres- sion forest. In CVPR, 2016. 12 Y .LIU, Y .W ANG, S.W ANG: A VCL

    2016

  21. [31]

    Saxena, M

    A. Saxena, M. Sun, and A. Y . Ng. Make3d: Learning 3d scene structure from a single still image. IEEE Transactions on Pattern Analysis and Machine Intelligence , 31(5): 824–840, 2009

    2009

  22. [32]

    Learning depth from single monocular images

    Ashutosh Saxena, Sung H Chung, and Andrew Y Ng. Learning depth from single monocular images. In NIPS, pages 1161–1168, 2006

    2006

  23. [33]

    Photorealistic scene reconstruction by voxel col- oring

    Steven M Seitz and Charles R Dyer. Photorealistic scene reconstruction by voxel col- oring. International Journal of Computer Vision, 35(2):151–173, 1999

    1999

  24. [34]

    Indoor segmenta- tion and support inference from rgbd images

    Nathan Silberman, Derek Hoiem, Pushmeet Kohli, and Rob Fergus. Indoor segmenta- tion and support inference from rgbd images. In ECCV, 2012

    2012

  25. [35]

    Efficient large-scale multi-view stereo for ultra high-resolution image sets

    Engin Tola, Christoph Strecha, and Pascal Fua. Efficient large-scale multi-view stereo for ultra high-resolution image sets. Machine Vision and Applications, 23(5):903–920, 2012

    2012

  26. [36]

    Disentangled representation learning gan for pose-invariant face recognition

    Luan Tran, Xi Yin, and Xiaoming Liu. Disentangled representation learning gan for pose-invariant face recognition. In CVPR, 2017

    2017

  27. [37]

    Peng Wang, Xiaohui Shen, Zhe Lin, Scott Cohen, Brian Price, and Alan L. Yuille. Towards unified depth and semantic prediction from a single image. In CVPR, 2015

    2015

  28. [38]

    A-fast-rcnn: Hard positive generation via adversary for object detection

    Xiaolong Wang, Abhinav Shrivastava, and Abhinav Gupta. A-fast-rcnn: Hard positive generation via adversary for object detection. In CVPR, 2017

    2017

  29. [39]

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

    Junyuan Xie, Ross Girshick, and Ali Farhadi. Deep3d: Fully automatic 2d-to-3d video conversion with deep convolutional neural networks. In ECCV, 2016

    2016

  30. [40]

    Multi-scale continuous crfs as sequential deep networks for monocular depth estimation

    Dan Xu, Elisa Ricci, Wanli Ouyang, Xiaogang Wang, and Nicu Sebe. Multi-scale continuous crfs as sequential deep networks for monocular depth estimation. InCVPR, 2017

    2017

  31. [41]

    Mvsnet: Depth inference for unstructured multi-view stereo

    Yao Yao, Zixin Luo, Shiwei Li, Tian Fang, and Long Quan. Mvsnet: Depth inference for unstructured multi-view stereo. In Proceedings of the European Conference on Computer Vision (ECCV), pages 767–783, 2018

    2018

  32. [42]

    Unsupervised learning of depth and ego-motion from video

    Tinghui Zhou, Matthew Brown, Noah Snavely, and David G Lowe. Unsupervised learning of depth and ego-motion from video. In CVPR, 2017

    2017

  33. [43]

    Generative visual manipulation on the natural image manifold

    Jun-Yan Zhu, Philipp Krähenbühl, Eli Shechtman, and Alexei A Efros. Generative visual manipulation on the natural image manifold. In ECCV, 2016

    2016

  34. [44]

    Indoor scene structure analysis for single image depth estimation

    Wei Zhuo, Mathieu Salzmann, Xuming He, and Miaomiao Liu. Indoor scene structure analysis for single image depth estimation. In CVPR, 2015

    2015

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.