Dedge-AGMNet:an effective stereo matching network optimized by depth edge auxiliary task

Weida Yang; Xindong Ai; Yong Xu; Yong Zhao; Zuliu Yang

REVIEW 3 major objections 4 minor 22 references

Dedge-AGMNet:an effective stereo matching network optimized by depth edge auxiliary task

T0 review · 3 major / 4 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 Depth-edge supervision plus granular 3D aggregation pushes stereo matching accuracy to reported state-of-the-art levels on SceneFlow and KITTI.

desk verdict A genuinely interesting depth-edge auxiliary idea and good SceneFlow numbers, but the KITTI SOTA claim is contradicted by the paper's own leaderboard results and the granular-convolution parameter math is wrong. read the letter →

arxiv 1908.09346 v4 pith:7ZELE52D submitted 2019-08-25 cs.CV eess.IV

Weida Yang , Xindong Ai , Zuliu Yang , Yong Xu , Yong Zhao This is my paper · ORCID

classification cs.CVeess.IV

keywords stereomatchingdisparityestimationdepthedgemulti-tasklearninggranularconvolutionatrousmulti-scaleaggregationcostvolume

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 claims that stereo depth estimation improves when the network is guided by a depth edge auxiliary task and aggregates cost volumes with a granular multi-scale module. The authors create depth edge ground truth by combining semantic and instance boundaries, share the feature extractor between the disparity and edge branches, and inject edge features through a modified spatial pyramid pooling stage called Dedge-SPP. They also introduce the AGM module, which applies granular convolution in 3D with parallel atrous branches to widen the receptive field while using fewer parameters. In their experiments, the full network reports the top published end-point-error on SceneFlow and fourth place on the KITTI 2012 and KITTI 2015 stereo test sets. If these results hold, the recipe provides a concrete way to improve disparity estimation in ambiguous boundary regions without requiring an edge network at inference time.

What carries the argument

The load-bearing mechanisms are the depth edge auxiliary branch and the AGM module. Depth edges are binary maps generated by overlaying instance boundaries on semantic boundaries, so overlapping objects of the same class remain separated; the branch shares weights with the main feature extractor and contributes a binary cross-entropy loss plus a disparity-gradient consistency loss, and Dedge-SPP concatenates edge features into the spatial pyramid pooling stage. The AGM module is built from granular convolution, where input channels are split into G groups and each group's output is fed into the next group's filter, giving multiple effective receptive fields; in AGM this is extended to 3D convolution and run as four parallel atrous branches inside an hourglass-style block. The machinery's purpose is to supply geometric boundary cues that suppress non-depth edge noise while aggregating multi-scale cost-volume context with claimed savings in parameters and computation.

What would settle it

Evaluate Eq. (5) at the channel and kernel dimensions the deployed AGM actually uses (for example C=32, kernel size 3, G=4) and count the trainable weights of one granular 3D convolution directly; if the measured ratio to a standard 3D convolution is not about 1/G, the paper's efficiency claim is false. A behavioral check: replace AGM with a standard 3D convolution matched to the same parameter budget and rerun the SceneFlow ablation; equal-or-better EPE from the standard convolution would show the granular structure is not the source of the improvement.

Watch

Extended reading notes

Core claim

The paper's central discovery, as the authors state it, is that depth edges—binary boundaries obtained by mining semantic and instance labels together—carry the supervision needed to resolve ill-posed stereo regions, and that an atrous granular multi-scale 3D module (AGM) is an effective and parameter-light aggregator. The depth edge branch shares its feature extractor with the disparity branch through hard parameter sharing, and Dedge-SPP embeds the edge features into the pyramid pooling stage. The AGM module repurposes granular convolution from 2D to 3D cost-volume aggregation, arranging four dilated granular branches in parallel to capture multi-scale context. The authors report that this combination reduces SceneFlow end-point-error by 27.0% relative to the pyramid stereo matching baseline, lowers KITTI 2015 validation D1-all from 1.67% to 1.54%, and places the method fourth on the KITTI 2012 and 2015 online test benchmarks while ranking first on SceneFlow among the compared published methods.

Load-bearing premise

The load-bearing premise is that the AGM module's granular convolution genuinely uses about 1/G the parameters of a standard convolution at the channel and kernel sizes used in practice; if Eq. (5)'s parameter count does not hold for the actual network dimensions, the efficiency rationale for the module collapses.

Editorial extensions

If this is right

Depth edge supervision can be added to an existing stereo network at little extra inference cost, since the edge branch is active only during training and Dedge-SPP adds about 0.1M parameters.
The AGM module with four parallel atrous granular convolutions at dilation rates 1, 4, 8, and 16 outperforms both the hourglass baseline and other dilation combinations on SceneFlow end-point-error.
The full model reports a 27.0% SceneFlow EPE reduction and a 17.4% KITTI 2015 validation D1-all reduction relative to the pyramid stereo baseline.
On public online benchmarks, the method reports the top published SceneFlow EPE among the compared methods (0.520) and fourth place on both KITTI 2012 and KITTI 2015 test sets.
Pre-training on an additional urban driving dataset in combination with SceneFlow further lowers the KITTI 2015 validation error, suggesting driving-domain data reinforces the depth-edge cues.

Reading between the lines

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

Editorial inference: the same depth-edge ground-truth recipe could be used to retrofit other cost-volume stereo networks, including real-time ones, because the edge branch is removed at inference and only the shared feature extractor and Dedge-SPP remain.
Editorial inference: if the parameter-saving formula is corrected and verified, a natural next test is to sweep the number of granular groups or atrous rates and measure the accuracy-latency frontier on embedded hardware; the paper fixes G=K=4 without such a sweep.
Editorial inference: the depth edge label could be generated from predicted semantic and instance masks rather than ground truth, enabling the auxiliary task for unlabeled or weakly labeled driving datasets; the paper only creates labels where semantic and instance ground truth exist.
Editorial inference: the loss-weight experiment on KITTI 2015 shows a=0 is best during fine-tuning, suggesting the depth-edge loss mainly helps during pre-training; a testable extension is to schedule the edge loss weight to decay during fine-tuning.

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.

Desk Editor's Note

read the letter

The abstract promises state-of-the-art on SceneFlow, KITTI 2012, and KITTI 2015; the paper's own Section 4.5 says it ranks fourth on both KITTI leaderboards. That is not a minor exaggeration—it is an internal refutation of the central claim. The SceneFlow number (0.520 EPE) does look genuinely strong, but the multi-benchmark SOTA claim is simply false.

What is actually new is the depth edge label: the union of semantic and instance boundaries, used as auxiliary supervision and fused via Dedge-SPP. That is a reasonable idea, and the ablations (Table 1) show it helps on SceneFlow and slightly on KITTI validation. The AGM module—parallel 3D granular convolutions with dilation rates 1, 4, 8, 16—also appears to give a nice improvement over PSMNet in their setup. I credit these components.

The soft spots are load-bearing. First, the parameter-count derivation in Eq. 5 is algebraically wrong. For C_in = C_out = C, the formula they write gives roughly s^2(G-1) + C^2, not (1/G)C^2s^2. So the claim that granular convolution uses about 1/G the parameters is not established; the actual savings (if any) depend on the exact grouping and point-wise projection in a way the paper does not rigorously analyze. Second, Table 2 shows the best KITTI validation result at a=0, i.e., with the depth edge loss term disabled. That undercuts the causal story for the auxiliary loss, even if the shared feature extraction and Dedge-SPP still help. Third, there are presentation issues: Adam beta1=0.09 looks like a typo, and the writing in places is hard to follow.

The paper is not a waste of time—the depth edge construction and the multi-scale aggregation idea could be worth pursuing. But as written, the claims are overstated and the math does not check out. If this lands on my desk, I would reject it and point the authors to fix the SOTA framing, redo the parameter analysis carefully, and clarify the role of the loss weight. It might become a solid workshop or second-tier paper after that.

Referee Report

3 major / 4 minor

Summary. The paper proposes Dedge-AGMNet, an end-to-end stereo matching network with two main contributions: a depth-edge auxiliary branch that shares features with the disparity branch and feeds them into a modified Dedge-SPP, and an atrous granular multi-scale (AGM) module built from granular convolutions intended to capture multi-scale context with fewer parameters than standard 3D convolutions. The authors generate depth-edge labels from semantic and instance segmentation maps, train on SceneFlow, Cityscapes, and KITTI, and report results on SceneFlow and the KITTI stereo benchmarks. The abstract and conclusion claim state-of-the-art performance on SceneFlow, KITTI 2012, and KITTI 2015, while the body reports first place on SceneFlow and fourth place on both KITTI leaderboards.

Significance. If the claimed contributions held, the paper would offer a practical multi-task architecture for stereo matching with a cheaper multi-scale aggregator and a principled depth-edge auxiliary task. The paper does provide useful experimental material: ablation studies on the depth-edge branch, the Dedge-SPP module, the AGM dilation configuration, and comparisons on public benchmarks. These ablations and the SceneFlow first-place result are concrete strengths. However, the central state-of-the-art claim is contradicted by the paper's own KITTI leaderboard results, and the parameter-count derivation that motivates the AGM module is algebraically incorrect. Because these two issues affect the paper's main claims, the contribution is not established as presented.

major comments (3)

[Abstract and Section 4.5] The abstract and conclusion state that Dedge-AGMNet 'outperforms other stereo matching networks and advances state-of-the-art performances on the Sceneflow, KITTI 2012 and KITTI 2015 benchmark datasets,' but Section 4.5 reports that the KITTI 2015 D1-all of 1.85% 'ranks in the fourth place' and that the KITTI 2012 result 'ranks fourth, too,' as shown in Tables 5 and 4. Fourth place on a public leaderboard is not state-of-the-art; at least three published methods have lower error on each benchmark. The multi-benchmark SOTA claim is therefore internally refuted and must be corrected, but the correction would remove the paper's headline contribution.
[Eq. (5), Section 3.3.2] The derivation that granular convolution has approximately 1/G times the parameters of standard convolution is algebraically wrong. With Cin=Cout=C and kernel size s, the cascade of G group convolutions, each with weights in R^{(C/G)×(C/G)×s×s}, has G·(C/G)^2·s^2 = C^2 s^2 / G parameters before the pointwise convolution, plus C^2 for the pointwise term. The expression in Eq. (5) instead contains a factor (G−1)/G^2 and treats the pointwise term as negligible; neither the factor nor the neglect is justified. For C=32, s=3, G=4, the actual ratio to a standard convolution is (9/4 + 1)/9 ≈ 0.36, not 1/4. Since the AGM module's design rationale and the paper's efficiency claim rest on this equation, the claim that granular convolution captures multi-scale information with fewer parameters is not supported.
[Table 2 and Section 4.3] The paper attributes part of the accuracy gain to the depth-edge auxiliary task, but Table 2 reports that on the KITTI 2015 validation set the best D1-all is obtained at a=0, which disables the depth-edge loss term Ledge in Eq. (8). Under that setting the auxiliary task has no direct loss contribution, so the experiment does not demonstrate that the depth-edge loss drives the reported fine-tuned improvement. This weakens a load-bearing causal claim of the paper and needs to be addressed explicitly, either by reporting results for a=0 in the ablation table or by explaining which other mechanism (e.g., Dedge-SPP) carries the benefit.

minor comments (4)

[Section 3.3.3] The running-latency relation 'RLparallel = G−1 = 1/K × Rcascade' is dimensionally inconsistent: the left side is presented as a ratio or count, while the right side includes 1/K times a latency; please clarify the definitions of K and the intended units.
[Table 5] In the row for EdgeStereo[17], the Non-Occluded D1-all value is reported as 3.00 while the All D1-all is 2.16; this is physically impossible, since non-occluded error cannot exceed the all-pixel error. This likely is a typographical error but should be corrected.
[Throughout] The dataset name is written inconsistently as 'Sceneflow', 'Sceneﬂow', and 'SceneFlow'; please use one spelling consistently.
[Section 3.2.2] The shared-concatenation notation '{F5(1),F,F5(2),F,...,F5(K),F}' is difficult to parse; a formal definition of the concatenation order would improve clarity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: this is an empirical stereo-matching systems paper; its benchmark results are supervised evaluations, and none of its architectural claims reduces to its inputs by construction.

full rationale

Dedge-AGMNet is an empirical systems paper rather than a first-principles derivation. The depth-edge labels are generated from semantic and instance ground truth, the networks are trained with supervised disparity and edge losses, and the reported SceneFlow/KITTI numbers are standard test-set evaluations after conventional fine-tuning. Hyperparameters such as a, gamma, and the dilation rates are chosen on validation sets, which is normal practice and does not constitute fitting a parameter and then renaming it a prediction. The granular-convolution module is explicitly adapted from Res2Net, and the parameter-count claim in Eq. 5 is an algebraic efficiency argument, not a self-referential definition; even if that algebra is doubtful, that is a correctness risk rather than circularity. There are no load-bearing self-citations: the cited prior works (PSMNet, GwcNet, EdgeStereo, CASENet, Res2Net) are external baselines and building blocks, not the authors' own unverified claims. The internal contradiction that the Abstract claims state-of-the-art on KITTI 2012/2015 while Section 4.5 reports fourth place on both leaderboards is a substantive accuracy/consistency problem, but it is not circularity because the reported ranks are externally determined benchmark outputs, not quantities defined in terms of the paper's own assumptions. Consequently, no circular step can be exhibited, and the appropriate score is 0.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

No new physical or architectural entity is postulated. The depth edge map is a data label derived from existing semantic and instance annotations, not an independent entity with external falsifiable predictions.

free parameters (5)

gamma = 0.5
Intensity of edge-gradient coupling in L_edgedisp (Eq 3), fixed in training, not swept.
loss weight a = 0 (best on KITTI 2015)
Balance between L_edge and L_edgedisp in Eq 8; Table 2 sweeps 0 to 1 on validation and chooses 0, meaning L_edge is disabled.
dilation rates = 1, 4, 8, 16
Selected by validation sweep in Table 3; other sets tested include 1,2,4,8 and 1,4,16,32.
G = K = 4
Number of granular groups and parallel branches; set by hand, stated in Section 3.3.3.
lambda1, lambda2, lambda3 = 0.5, 0.7, 1.0
Weights for the three disparity output losses in Eq 7, chosen by hand.

assumptions (3)

domain assumption Depth edge, defined as union of semantic and instance boundaries, reliably encodes disparity discontinuities.
Section 3.2.1 asserts this as a 'general fact' and supports it with qualitative figure 4 rather than quantitative evidence.
domain assumption Hard parameter sharing between the depth edge branch and stereo branch transfers useful geometric constraints without harming the main task.
Section 3.2.3 implements weight sharing; ablation in Table 1 supports it but the mechanism is not proven.
ad hoc to paper The simplification Ngranular approximately equals (1/G) times Nstandard in Eq 5 is mathematically valid.
This is the paper's own derivation; with C input/output channels and kernel s, the correct result includes a C^2 term, so the stated 1/G reduction does not hold for typical channel counts.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Dedge-AGMNet:an effective stereo matching network optimized by depth edge auxiliary task." pith.science (2026). https://pith.science/paper/7ZELE52D

@misc{pith2026190809346,
  author       = {Pith},
  title        = {Pith review of: Dedge-AGMNet:an effective stereo matching network optimized by depth edge auxiliary task},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7ZELE52D}},
  note         = {Machine review of arXiv:1908.09346}
}

read the original abstract

To improve the performance in ill-posed regions, this paper proposes an atrous granular multi-scale network based on depth edge subnetwork(Dedge-AGMNet). According to a general fact, the depth edge is the binary semantic edge of instance-sensitive. This paper innovatively generates the depth edge ground-truth by mining the semantic and instance dataset simultaneously. To incorporate the depth edge cues efficiently, our network employs the hard parameter sharing mechanism for the stereo matching branch and depth edge branch. The network modifies SPP to Dedge-SPP, which fuses the depth edge features to the disparity estimation network. The granular convolution is extracted and extends to 3D architecture. Then we design the AGM module to build a more suitable structure. This module could capture the multi-scale receptive field with fewer parameters. Integrating the ranks of different stereo datasets, our network outperforms other stereo matching networks and advances state-of-the-art performances on the Sceneflow, KITTI 2012 and KITTI 2015 benchmark datasets.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

22 extracted references · 19 canonical work pages

[1]

11786–11795, (2019)

Rohan Chabra, Julian Straub, Christopher Sweeney, Richard New- combe, and Henry Fuchs, ‘Stereodrnet: Dilated residual stereonet’, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 11786–11795, (2019)

work page 2019
[2]

5410–5418, (2018)

Jia-Ren Chang and Yong-Sheng Chen, ‘Pyramid stereo matching net- work’, inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 5410–5418, (2018)

work page 2018
[3]

Liang-Chieh Chen, George Papandreou, Florian Schroff, and Hartwig Adam, ‘Rethinking atrous convolution for semantic image segmenta- tion’,arXiv preprint arXiv:1706.05587, (2017)

arXiv 2017
[4]

Xinjing Cheng, Peng Wang, and Ruigang Yang, ‘Learning depth with convolutional spatial propagation network’, arXiv preprint arXiv:1810.02695, (2018)

work page Pith review arXiv 2018
[5]

Xianzhi Du, Mostafa El-Khamy, and Jungwon Lee, ‘Amnet: Deep atrous multiscale stereo disparity estimation networks’, arXiv preprint arXiv:1904.09099, (2019)

arXiv 2019
[6]

Shang-Hua Gao, Ming-Ming Cheng, Kai Zhao, Xin-Yu Zhang, Ming- Hsuan Yang, and Philip Torr, ‘Res2net: A new multi-scale backbone architecture’,arXiv preprint arXiv:1904.01169, (2019)

arXiv 2019
[7]

Xiaoyang Guo, Kai Yang, Wukui Yang, Xiaogang Wang, and Hong- sheng Li, ‘Group-wise correlation stereo network’, 3273–3282, (2019)

work page 2019
[8]

Heiko Hirschmuller, ‘Accurate and efﬁcient stereo processing by semi- global matching and mutual information’, in 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), volume 2, pp. 807–814. IEEE, (2005)

work page 2005

Show all 22 references

[9]

4700–4708, (2017)

Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q Wein- berger, ‘Densely connected convolutional networks’, inProceedings of the IEEE conference on computer vision and pattern recognition , pp. 4700–4708, (2017)

2017
[10]

66–75, (2017)

Alex Kendall, Hayk Martirosyan, Saumitro Dasgupta, Peter Henry, Ryan Kennedy, Abraham Bachrach, and Adam Bry, ‘End-to-end learn- ing of geometry and context for deep stereo regression’, inProceedings of the IEEE International Conference on Computer Vision , pp. 66–75, (2017)

2017
[11]

3431–3440, (2015)

Jonathan Long, Evan Shelhamer, and Trevor Darrell, ‘Fully convolu- tional networks for semantic segmentation’, inProceedings of the IEEE conference on computer vision and pattern recognition, pp. 3431–3440, (2015)

2015
[12]

5695–5703, (2016)

Wenjie Luo, Alexander G Schwing, and Raquel Urtasun, ‘Efﬁcient deep learning for stereo matching’, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 5695–5703, (2016)

2016
[13]

4040–4048, (2016)

Nikolaus Mayer, Eddy Ilg, Philip Hausser, Philipp Fischer, Daniel Cre- mers, Alexey Dosovitskiy, and Thomas Brox, ‘A large dataset to train convolutional networks for disparity, optical ﬂow, and scene ﬂow esti- mation’, in Proceedings of the IEEE Conference on Computer Vision ...

2016
[14]

3260–3269, (2017)

Thomas Schops, Johannes L Schonberger, Silvano Galliani, Torsten Sattler, Konrad Schindler, Marc Pollefeys, and Andreas Geiger, ‘A multi-view stereo benchmark with high-resolution images and multi- camera videos’, in Proceedings of the IEEE Conference on Computer Vision and Pa...

2017
[15]

231–240, (2017)

Akihito Seki and Marc Pollefeys, ‘Sgm-nets: Semi-global matching with neural networks’, inProceedings of the IEEE Conference on Com- puter Vision and Pattern Recognition, pp. 231–240, (2017)

2017
[16]

Xiao Song, Xu Zhao, Liangji Fang, and Hanwen Hu, ‘Edgestereo: An effective multi-task learning network for stereo matching and edge de- tection’,arXiv preprint arXiv:1903.01700, (2019)

2019 arXiv
[17]

Xiao Song, Xu Zhao, Hanwen Hu, and Liangji Fang, ‘Edgestereo: A context integrated residual pyramid network for stereo matching’, in Asian Conference on Computer Vision, pp. 20–35. Springer, (2018)

2018
[18]

1395–1403, (2015)

Saining Xie and Zhuowen Tu, ‘Holistically-nested edge detection’, in Proceedings of the IEEE international conference on computer vision , pp. 1395–1403, (2015)

2015
[19]

636–651, (2018)

Guorun Yang, Hengshuang Zhao, Jianping Shi, Zhidong Deng, and Ji- aya Jia, ‘Segstereo: Exploiting semantic information for disparity esti- mation’, in Proceedings of the European Conference on Computer Vi- sion (ECCV), pp. 636–651, (2018)

2018
[20]

3684–3692, (2018)

Maoke Yang, Kun Yu, Chi Zhang, Zhiwei Li, and Kuiyuan Yang, ‘Denseaspp for semantic segmentation in street scenes’, inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pp. 3684–3692, (2018)

2018
[21]

5964–5973, (2017)

Zhiding Yu, Chen Feng, Ming-Yu Liu, and Srikumar Ramalingam, ‘Casenet: Deep category-aware semantic edge detection’, in Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pp. 5964–5973, (2017)

2017
[22]

Jure Zbontar, Yann LeCun, et al., ‘Stereo matching by training a convo- lutional neural network to compare image patches.’,Journal of Machine Learning Research, 17(1-32), 2, (2016)

2016

Pith tools

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

[1] [1]

11786–11795, (2019)

Rohan Chabra, Julian Straub, Christopher Sweeney, Richard New- combe, and Henry Fuchs, ‘Stereodrnet: Dilated residual stereonet’, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 11786–11795, (2019)

work page 2019

[2] [2]

5410–5418, (2018)

Jia-Ren Chang and Yong-Sheng Chen, ‘Pyramid stereo matching net- work’, inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 5410–5418, (2018)

work page 2018

[3] [3]

Liang-Chieh Chen, George Papandreou, Florian Schroff, and Hartwig Adam, ‘Rethinking atrous convolution for semantic image segmenta- tion’,arXiv preprint arXiv:1706.05587, (2017)

arXiv 2017

[4] [4]

Xinjing Cheng, Peng Wang, and Ruigang Yang, ‘Learning depth with convolutional spatial propagation network’, arXiv preprint arXiv:1810.02695, (2018)

work page Pith review arXiv 2018

[5] [5]

Xianzhi Du, Mostafa El-Khamy, and Jungwon Lee, ‘Amnet: Deep atrous multiscale stereo disparity estimation networks’, arXiv preprint arXiv:1904.09099, (2019)

arXiv 2019

[6] [6]

Shang-Hua Gao, Ming-Ming Cheng, Kai Zhao, Xin-Yu Zhang, Ming- Hsuan Yang, and Philip Torr, ‘Res2net: A new multi-scale backbone architecture’,arXiv preprint arXiv:1904.01169, (2019)

arXiv 2019

[7] [7]

Xiaoyang Guo, Kai Yang, Wukui Yang, Xiaogang Wang, and Hong- sheng Li, ‘Group-wise correlation stereo network’, 3273–3282, (2019)

work page 2019

[8] [8]

Heiko Hirschmuller, ‘Accurate and efﬁcient stereo processing by semi- global matching and mutual information’, in 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05), volume 2, pp. 807–814. IEEE, (2005)

work page 2005