Structured Coupled Generative Adversarial Networks for Unsupervised Monocular Depth Estimation

Andrea Pilzer; Dan Xu; Mihai Marian Puscas; Nicu Sebe

REVIEW 5 major objections 5 minor 33 references

Structured Coupled Generative Adversarial Networks for Unsupervised Monocular Depth Estimation

T0 review · 5 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 A dual GAN fused by a deep CRF reaches 0.1354 mean relative error on KITTI unsupervised depth.

desk verdict Solid empirical architecture paper whose central claim—a CRF that couples generator and discriminator—is not actually in the equations; the paper runs the same CRF twice and sums gradients, so the main novelty is unsupported as written. read the letter →

arxiv 1908.05794 v1 pith:EMY6T4FA submitted 2019-08-15 cs.CV eess.IV

Mihai Marian Puscas , Dan Xu , Andrea Pilzer , Nicu Sebe This is my paper · ORCID

classification cs.CVeess.IV

keywords monoculardepthestimationunsupervisedlearninggenerativeadversarialnetworksconditionalrandomfieldsdisparitystereoimagepairsdeepstructuredprediction

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

This paper proposes an unsupervised way to estimate depth from a single image without ground-truth depth labels, using only stereo image pairs during training. Its central claim is that coupling two GAN branches through a deep conditional random field improves disparity estimation, because the CRF fuses both the two generated disparity maps and the two discriminators' error maps in a structured, spatial way. A sympathetic reader should care because the method reports the best numbers among the compared unsupervised methods that train on image pairs only, and still needs just one image at test time. The argument is that generative and discriminative maps carry complementary structural information, and that fusing them with a learnable CRF is what produces the gain.

What carries the argument

The central object is the deep CRF coupling model: a continuous conditional random field whose hidden disparity field $d_r$ is refined by mean-field inference implemented as differentiable neural-network blocks (NMF). Its energy combines two unary terms that tie the hidden field to the two generator disparity maps $d_{ra}$ and $d_{rh}$, plus Gaussian appearance and smoothness kernels over a local $15 \times 15$ receptive field that let nearby pixels exchange messages; the same CRF is applied to the two discriminators' pixel-level score maps to produce a fused adversarial score map. The mean-field update, $d_i^r = \frac{\alpha_1 d_{ra,i} + \alpha_2 d_{rh,i} + \sum_l \sum_{j\neq i} k_l(f_i,f_j) d_j^r}{\alpha_1 + \alpha_2 + \sum_l \sum_{j\neq i} k_l(f_i,f_j)}$, is what couples the branches: gradients from the fused adversarial loss flow into both generators and both discriminators, so the CRF is the mechanism that turns 'generate and discriminate' into 'generate, discriminate, and refine together'.

What would settle it

Measure, on KITTI, the per-pixel correlation between the photometric reconstruction error of each synthesized image and the corresponding discriminator fake-score map; if the correlation is near zero or the high-error regions do not overlap, the premise behind the CRF coupling is unsupported. A direct ablation would do the same job: replace the CRF coupling with a simple channel-concatenation or weighted average of the two disparity maps and two score maps; if the CRF version does not beat that fusion on the reported metrics, the structured coupling is not the source of the gain.

Watch

Extended reading notes

Core claim

The paper's core discovery is that explicit, structured coupling of a dual GAN improves unsupervised monocular depth estimation. Two generators produce two complementary disparity maps from the two stereo views, a pixel-level discriminator scores each synthesized image, and then one deep CRF, implemented as a differentiable mean-field network, refines and fuses the two disparity maps and the two adversarial score maps in the same model. The CRF's unary terms pull the hidden disparity toward both generator outputs, while Gaussian appearance and smoothness kernels propagate information between nearby pixels; gradients from the fused adversarial score map flow back into both discriminators and generators. With this joint coupling, the CRF-DGAN reaches a mean relative error of 0.1354 on KITTI at the 80-meter cap and 0.1283 at 50 meters, outperforming the compared unsupervised methods that use only image pairs during training.

Load-bearing premise

The whole method rests on the assumption that a pixel's generative error (how badly image reconstruction fails there) and its discriminative error (how fake the discriminator judges that region) line up spatially, so that fusing the two maps with a CRF has real structure to exploit; if those maps are uncorrelated or misaligned, the coupled CRF has no signal to work with.

Editorial extensions

If this is right

On KITTI, the full model reports a mean relative error of 0.1354 at the 80-meter cap and 0.1283 at 50 meters, with 82.8 percent of pixels within the 1.25 threshold at 80 meters, beating all compared unsupervised methods that train on image pairs only.
The system needs stereo image pairs only during training; at test time a hallucination subnetwork produces a disparity map from a single image, so the method applies where only a monocular camera is available.
Coupling both discriminators and generators with the CRF improves over coupling discriminators alone in the paper's ablations, showing that mutual constraints between generator and discriminator contribute beyond the dual-branch structure.
Because the same CRF machinery is applied to adversarial score maps, the scheme is a template for other GAN-based dense prediction tasks with rich spatial structure, as the paper itself notes.

Reading between the lines

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

The CRF coupling is a general module, not a depth-specific one; it should transfer to other dense prediction tasks with two complementary branches and pixel-level discriminators, such as optical flow, scene flow, or surface normal estimation, though the paper does not test this.
A quantitative test of the paper's central premise is missing: correlating per-pixel generative error with discriminator score would tell whether the CRF is exploiting real structure or simply acting as a learned smoother.
Because the hallucination network is trained with an L1 loss between the two disparity maps, the monocular test-time path is only as good as that approximation; one could push further by adding the CRF-fused disparity as supervision for the hallucinator.
The local 15-by-15 message-passing window means the CRF's structured prior is largely local; a fully connected or multi-scale version could change the gains, separating the benefit of the CRF model from the benefit of a larger receptive field.

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.

Desk Editor's Note

read the letter

Quick take: the paper is worth reading as an incremental architecture study, but the main advertised mechanism—a CRF that jointly couples generator and discriminator—is not present in the equations. What the paper actually does is run the same continuous CRF twice, once on the disparity maps and once on the discriminator score maps, and sum the gradients. That is shared-weight post-processing plus gradient averaging, not a structured model of the co-occurrence between depth errors and adversarial errors.

What is genuinely there: a dual-GAN with a hallucination branch is a sensible design for monocular inference, and the continuous CRF with mean-field updates is implemented as a differentiable network, which is a real engineering effort. The evaluation is reasonably broad (KITTI, Cityscapes, Make3D), the training protocol is standard, and the reported numbers, if they hold up, are competitive for 2019. The ablations show incremental gains, and the paper does not fit to test sets. Self-citation of the authors' prior CRF and GAN work is normal here and not a problem.

The soft spots are substantial. Eq. 3 writes an energy over a single hidden disparity variable; there is no adversarial score variable in the model. Eq. 4 is malformed, and Eq. 5 drops the beta_l kernel weights that Eq. 3 introduced, reducing the written update to a fixed bilateral smoother and contradicting the 'four parameters' claim. Eq. 2 has swapped left/right subscripts or worse. Table 1 contains a physically impossible negative RMS log value (-0.773) in the baseline. That is likely a typo, but it makes you squint at the rest of the table. Because the mechanism is not modeled, the 1.2-point rel gain in the ablation cannot be attributed to the claimed structured coupling; it could be extra capacity or regularization. There is also no comparison to the contemporary Monodepth2 line of work, which matters for a state-of-the-art claim. Finally, the abstract says code is available while the intro says it will be released upon acceptance—one of those is wrong.

Who is this for? Someone tracking self-supervised depth estimation and graphical-model refinement will get a snapshot of a 2019 approach that didn't catch on. A referee should engage, but the authors need to rewrite the method section so the equations actually describe the model, correct the tables, and release the code/checkpoints before the claims can be verified. With that, the empirical core might stand; as is, the paper's central novelty is not supported.

Referee Report

5 major / 5 minor

Summary. The manuscript proposes CRF-DGAN, an end-to-end unsupervised framework for monocular depth estimation. Two generator/discriminator pairs operating on stereo images synthesize right-view images; a hallucination sub-network converts the left-view disparity to an approximate right-view disparity so that only a single image is needed at test time; and a deep continuous CRF, implemented as differentiable mean-field iterations, is applied to refine both the generator disparity maps and the pixel-level discriminator score maps. The total loss combines reconstruction, hallucination, adversarial, and CRF reconstruction terms. Experiments on KITTI, Cityscapes, and Make3D report state-of-the-art performance among unsupervised methods using only image pairs, and an ablation study attributes gains to each component of the model.

Significance. If the reported results are correct, the paper makes a solid contribution to unsupervised monocular depth estimation by demonstrating that CRF-based structured refinement can be combined with adversarial training in a dual-GAN architecture. Strengths include the breadth of experiments (three datasets, many baselines), the component-wise ablation, and the stated intention to release code. The central novelty—the 'structured coupling' of generators and discriminators—is appealing, but the manuscript as written does not accurately instantiate this idea in the equations, and the reported ablation gains cannot be unambiguously attributed to it. These issues are correctable in revision, but they are load-bearing for the paper's main claim.

major comments (5)

[§3.2, Eq. (4)] Equation (4) is malformed: it contains the undefined term 'd_i^r]', and the pairwise summation over j is missing from the exponential, so the mean-field Gaussian update cannot be verified as following from Eq. (3). Because the NMF implementation and its backpropagation rely on this update, please provide a corrected derivation and state the assumed form of the variational distribution.
[§3.2, Eq. (5)] Equation (5) drops the beta_l kernel weights introduced in Eq. (3); with the update as written, the message-passing terms are unweighted bilateral filters, and the claimed 'four parameters' (alpha_1, alpha_2, beta_1, beta_2) reduce to only two learnable weights. Please correct Eq. (5) to include beta_l in both numerator and denominator and clarify exactly which parameters are optimized during training.
[§3.2, 'Joint Coupling of the Generator and Discriminator'] The described coupling procedure runs two separate forward computations of the same CRF—one on disparity maps and one on adversarial score maps—and then sums gradients in a single backward pass. This is shared-weight smoothing of two outputs, not a joint probabilistic model; Eq. (3) defines an energy over only the hidden disparity variable and contains no term linking it to a hidden adversarial-score variable. Consequently, the 1.2-point rel improvement in Table 1 (variant v versus iii) cannot be attributed to structured cross-modal coupling. Please either add an explicit coupling term to the CRF energy or revise the stated contribution to describe what is actually implemented.
[Table 1] The baseline row reports an RMS log error of -0.773, which is impossible for a root-mean-square quantity, and the row contains only six entries for the seven metric columns shown in the header. Please correct the table, verify all numerical values, and ensure the ablation comparisons cited in the text (e.g., the '0.779 to 0.813' accuracy gain) correspond to the correct rows.
[§3.1, Eq. (2)] The second term of Eq. (2) compares the synthesized left image \tilde{I}_l with the right image I_r, and the definitions of \tilde{I}_l and \tilde{I}_r appear to be swapped relative to the disparity outputs of G_a and G_b described in the text. Please clarify the notation (d_l versus d_r, G_a versus G_b) and correct the reconstruction loss so that each synthesized image is compared with its corresponding real view.

minor comments (5)

[Section 1 and Fig. 7] The paper motivates the CRF coupling by asserting that depth prediction maps and discriminative error maps share spatially co-located structural information, but only qualitative examples are provided. A quantitative measurement of this correlation would considerably strengthen the motivation.
[§4.2, ablation discussion] The sentence 'we improve 4 points from 0.779 to 0.813' in the ablation discussion refers to a comparison between variants (ii) and (iv) in Table 1, not to a direct comparison of (v) and (iii); please make the compared rows explicit.
[Abstract and footnote] The code URL is given as 'https://github.com/mihaipuscas/ 3dv---coupled-crf-disparity' with a space in the middle; the correct link should be provided.
[§3.1, Eq. (1)] In Eq. (1), the third and fourth expectations are both with respect to p(I_n^r), while the second is with respect to p(I_n^l); given the described roles of G_a and G_b, the input distributions of the two generators should be stated consistently.
[§4.1 and §4.2] The number of mean-field iterations T and the values of the loss weights gamma_1, gamma_2, gamma_3 are not reported; these details are needed to reproduce the results.

Circularity Check

0 steps flagged · score 1.0 of 10

No load-bearing circularity: the depth prediction is anchored to external benchmarks and learned CRF parameters; self-citations are incremental, not definitional.

full rationale

The central depth prediction is trained with photometric, adversarial, hallucination, and CRF losses and evaluated against LiDAR ground truth on KITTI, Cityscapes, and Make3D. No parameter is fitted to the test set and then renamed as a prediction. Eq. 3 defines a standard continuous CRF energy with unary terms on the two generator disparity maps and pairwise RGB kernels; Eq. 5 is the corresponding mean-field update. The CRF weights are learned end-to-end on training data, and the final disparity map is an evaluated output, not a re-statement of the training target. The paper cites several works by the same authors (e.g., [19], [20], [28], [29], [30]), but the CRF equations are written out in the paper and the self-citations are incremental related work rather than an unverified uniqueness theorem. The manuscript does contain a genuine internal consistency problem: the claimed joint coupling of generator and discriminator is described as two separate CRF forward computations with collected gradients, and Eq. 5 contains no adversarial-score variable; Table 1 also reports an impossible negative RMS log value. These are correctness concerns, not circular reductions, so under the instruction to flag circularity only where Eq. X equals Eq. Y by construction or a fitted parameter is renamed as a prediction, no circular step is exhibited. The honest finding is therefore a low score with no circular steps.

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

The central claim rests on standard unsupervised depth estimation assumptions (photometric consistency, differentiable warping) and on a paper-specific assumption that discriminator score maps are spatially correlated with depth errors. Hyperparameters such as loss weights and CRF kernel size are chosen by hand and not fully reported, which makes exact replication harder.

free parameters (3)

loss weights gamma1, gamma2, gamma3 = not reported
Weights balancing reconstruction, hallucination, and adversarial losses in the total objective. The paper states a set of weights but does not provide values.
CRF kernel window size = 15x15
Local receptive field for the Gaussian convolution in message passing; chosen by hand and affects the structured refinement quality.
number of mean-field iterations T = not specified
The mean-field update is iterative, but the number of iterations used in the neural network implementation is not reported.

assumptions (4)

domain assumption Photometric warping assumes Lambertian surfaces and known stereo calibration
The reconstruction loss compares synthesized and real images; violations such as specularities and occlusions reduce supervision quality. Invoked in Section 3.1.
standard math Differentiable bilinear warping is a valid proxy for view synthesis
The warping function fw is used to synthesize images from disparity maps; this is standard in the depth estimation literature.
standard math Mean-field approximation of the continuous CRF yields a tractable and effective inference
The update in Eq. 5 is derived from mean-field theory and is an approximation to exact inference over the fully connected graph.
domain assumption Pixel-level discriminator outputs carry spatial structure correlated with depth errors
The coupling of generator and discriminator outputs via CRF relies on this correlation, asserted in Section 1 but not quantitatively verified.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Structured Coupled Generative Adversarial Networks for Unsupervised Monocular Depth Estimation." pith.science (2026). https://pith.science/paper/EMY6T4FA

@misc{pith2026190805794,
  author       = {Pith},
  title        = {Pith review of: Structured Coupled Generative Adversarial Networks for Unsupervised Monocular Depth Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EMY6T4FA}},
  note         = {Machine review of arXiv:1908.05794}
}

read the original abstract

Inspired by the success of adversarial learning, we propose a new end-to-end unsupervised deep learning framework for monocular depth estimation consisting of two Generative Adversarial Networks (GAN), deeply coupled with a structured Conditional Random Field (CRF) model. The two GANs aim at generating distinct and complementary disparity maps and at improving the generation quality via exploiting the adversarial learning strategy. The deep CRF coupling model is proposed to fuse the generative and discriminative outputs from the dual GAN nets. As such, the model implicitly constructs mutual constraints on the two network branches and between the generator and discriminator. This facilitates the optimization of the whole network for better disparity generation. Extensive experiments on the KITTI, Cityscapes, and Make3D datasets clearly demonstrate the effectiveness of the proposed approach and show superior performance compared to state of the art methods. The code and models are available at https://github.com/mihaipuscas/ 3dv---coupled-crf-disparity.

Figures

Figures reproduced from arXiv: 1908.05794 by the authors.

**Figure 2.** Framework overview of the proposed CRF-DGAN for unsupervised monocular depth estimation. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Examples of the structured output of the disparity maps [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Examples of depth prediction results on the KITTI raw dataset. Qualitative comparison with other depth estimation methods on [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: Qualitative comparison of different variants of the proposed CRF-DGAN model on the Cityscapes dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 7.** Figure 7: Examples of structured outputs of the real and the fake [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 32 canonical work pages

[1]

Adams, J

A. Adams, J. Baek, and M. A. Davis. Fast high-dimensional ﬁltering using the permutohedral lattice. In Computer Graphics F orum, 2010

work page 2010
[2]

A. Behl, O. H. Jafari, S. K. Mustikovela, H. A. Alhaija, C. Rother, and A. Geiger. Bounding boxes, segmentations and object coordinates: How important is recognition for 3d scene ﬂow estimation in autonomous driving scenarios? In CVPR, 2017

work page 2017
[3]

Cordts, M

M. Cordts, M. Omran, S. Ramos, T. Rehfeld, M. Enzweiler, R. Benenson, U. Franke, S. Roth, and B. Schiele. The cityscapes dataset for semantic urban scene understanding. In CVPR, 2016

2016
[4]

Eigen and R

D. Eigen and R. Fergus. Predicting depth, surface normals and semantic labels with a common multi-scale convolu- tional architecture. In ICCV, 2015

work page 2015
[5]

Eigen, C

D. Eigen, C. Puhrsch, and R. Fergus. Depth map prediction from a single image using a multi-scale deep network. In NIPS, 2014

work page 2014
[6]

Y . Gan, X. Xu, W. Sun, and L. Lin. Monocular depth estima- tion with afﬁnity, vertical pooling, and label enhancement. In ECCV, 2018

work page 2018
[7]

R. Garg, G. Carneiro, and I. Reid. Unsupervised cnn for single view depth estimation: Geometry to the rescue. In ECCV, 2016

work page 2016
[8]

Geiger, P

A. Geiger, P. Lenz, C. Stiller, and R. Urtasun. Vision meets robotics: The kitti dataset. IJRR, 2013

work page 2013

Show all 33 references

[9]

Godard, O

C. Godard, O. Mac Aodha, and G. J. Brostow. Unsuper- vised monocular depth estimation with left-right consistency. CVPR, 2017

2017
[10]

Goodfellow, J

I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio. Gen- erative adversarial nets. In NIPS, 2014

2014
[11]

X. Guo, H. Li, S. Yi, J. Ren, and X. Wang. Learning monoc- ular depth by distilling cross-domain stereo networks. In ECCV, 2018

2018
[12]

Kr ¨ahenb¨uhl and V

P. Kr ¨ahenb¨uhl and V . Koltun. Efﬁcient inference in fully connected crfs with gaussian edge potentials. NIPS, 2011

2011
[13]

J. N. Kundu, P. K. Uppala, A. Pahuja, and R. V . Babu. Adadepth: Unsupervised content congruent adaptation for depth estimation. In CVPR, 2018

2018
[14]

Kuznietsov, J

Y . Kuznietsov, J. St ¨uckler, and B. Leibe. Semi-supervised deep learning for monocular depth map prediction. CVPR, 2017

2017
[15]

Laina, C

I. Laina, C. Rupprecht, V . Belagiannis, F. Tombari, and N. Navab. Deeper depth prediction with fully convolutional residual networks. arXiv preprint arXiv:1606.00373, 2016

2016 arXiv
[16]

F. Liu, C. Shen, G. Lin, and I. Reid. Learning depth from sin- gle monocular images using deep convolutional neural ﬁelds. TPAMI, 2016

2016
[17]

Mahjourian, M

R. Mahjourian, M. Wicke, and A. Angelova. Unsupervised learning of depth and ego-motion from monocular video us- ing 3d geometric constraints. In CVPR, 2018

2018
[18]

Mayer, E

N. Mayer, E. Ilg, P. Hausser, P. Fischer, D. Cremers, A. Dosovitskiy, and T. Brox. A large dataset to train convo- lutional networks for disparity, optical ﬂow, and scene ﬂow estimation. In CVPR, 2016

2016
[19]

Pilzer, S

A. Pilzer, S. Lathuiliere, N. Sebe, and E. Ricci. Reﬁne and distill: Exploiting cycle-inconsistency and knowledge dis- tillation for unsupervised monocular depth estimation. In CVPR, 2019

2019
[20]

Pilzer, D

A. Pilzer, D. Xu, M. Puscas, E. Ricci, and N. Sebe. Unsuper- vised adversarial depth estimation using cycled generative networks. In 3DV, 2018

2018
[21]

Ristovski, V

K. Ristovski, V . Radosavljevic, S. Vucetic, and Z. Obradovic. Continuous conditional random ﬁelds for efﬁcient regression in large fully connected graphs. In AAAI, 2013

2013
[22]

Saxena, S

A. Saxena, S. H. Chung, and A. Y . Ng. Learning depth from single monocular images. In NIPS, 2006

2006
[23]

Saxena, M

A. Saxena, M. Sun, and A. Y . Ng. Make3d: Learning 3d scene structure from a single still image. TPAMI, 2009

2009
[24]

Turan, Y

M. Turan, Y . Almalioglu, H. Araujo, E. Konukoglu, and M. Sitti. A non-rigid map fusion-based rgb-depth slam method for endoscopic capsule robots. arXiv preprint arXiv:1705.05444, 2017

2017 arXiv
[25]

C. Wang, J. M. Buenaposada, R. Zhu, and S. Lucey. Learn- ing depth from monocular videos using direct methods. In CVPR, 2018

2018
[26]

P. Wang, X. Shen, Z. Lin, S. Cohen, B. Price, and A. Yuille. Towards uniﬁed depth and semantic prediction from a single image. In CVPR, 2015

2015
[27]

Xie and Z

S. Xie and Z. Tu. Holistically-nested edge detection. In ICCV, 2015

2015
[28]

D. Xu, E. Ricci, W. Ouyang, X. Wang, and N. Sebe. Multi-scale continuous CRFs as sequential deep networks for monocular depth estimation. CVPR, 2017

2017
[29]

D. Xu, E. Ricci, W. Ouyang, X. Wang, and N. Sebe. Monoc- ular depth estimation using multi-scale continuous CRFs as sequential deep networks. TPAMI, 41(6):1426–1440, 2018

2018
[30]

D. Xu, W. Wang, H. Tang, H. Liu, N. Sebe, and E. Ricci. Structured attention guided convolutional neural ﬁelds for monocular depth estimation. In CVPR, 2018

2018
[31]

H. Zhan, R. Garg, C. S. Weerasekera, K. Li, H. Agarwal, and I. Reid. Unsupervised learning of monocular depth estima- tion and visual odometry with deep feature reconstruction. arXiv preprint arXiv:1803.03893, 2018

2018 arXiv
[32]

T. Zhou, M. Brown, N. Snavely, and D. G. Lowe. Unsuper- vised learning of depth and ego-motion from video. CVPR, 2017

2017
[33]

Y . Zou, Z. Luo, and J.-B. Huang. Df-net: Unsupervised joint learning of depth and ﬂow using cross-task consistency. In ECCV, 2018

2018

Pith tools

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

[1] [1]

Adams, J

A. Adams, J. Baek, and M. A. Davis. Fast high-dimensional ﬁltering using the permutohedral lattice. In Computer Graphics F orum, 2010

work page 2010

[2] [2]

A. Behl, O. H. Jafari, S. K. Mustikovela, H. A. Alhaija, C. Rother, and A. Geiger. Bounding boxes, segmentations and object coordinates: How important is recognition for 3d scene ﬂow estimation in autonomous driving scenarios? In CVPR, 2017

work page 2017

[3] [3]

Cordts, M

M. Cordts, M. Omran, S. Ramos, T. Rehfeld, M. Enzweiler, R. Benenson, U. Franke, S. Roth, and B. Schiele. The cityscapes dataset for semantic urban scene understanding. In CVPR, 2016

2016

[4] [4]

Eigen and R

D. Eigen and R. Fergus. Predicting depth, surface normals and semantic labels with a common multi-scale convolu- tional architecture. In ICCV, 2015

work page 2015

[5] [5]

Eigen, C

D. Eigen, C. Puhrsch, and R. Fergus. Depth map prediction from a single image using a multi-scale deep network. In NIPS, 2014

work page 2014

[6] [6]

Y . Gan, X. Xu, W. Sun, and L. Lin. Monocular depth estima- tion with afﬁnity, vertical pooling, and label enhancement. In ECCV, 2018

work page 2018

[7] [7]

R. Garg, G. Carneiro, and I. Reid. Unsupervised cnn for single view depth estimation: Geometry to the rescue. In ECCV, 2016

work page 2016

[8] [8]

Geiger, P

A. Geiger, P. Lenz, C. Stiller, and R. Urtasun. Vision meets robotics: The kitti dataset. IJRR, 2013

work page 2013