Pith. sign in

REVIEW 3 major objections 5 minor 65 references

Integrating Disparity Confidence Estimation into Relative Depth Prior-Guided Unsupervised Stereo Matching

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper claims that an unsupervised stereo matching framework, combining disparity confidence voting with two relative-depth-prior losses, reaches state-of-the-art accuracy on the KITTI Stereo benchmarks among all unsupervised methods.

desk verdict A solid incremental unsupervised-stereo paper whose apparent fatal flaw in Eq. (3) evaporates on close reading; the real caveats are a typo and the shared-encoder prior. read the letter →

arxiv 2508.01275 v1 pith:3HFKBB3L submitted 2025-08-02 cs.CV

Chuang-Wei Liu , Mingjian Sun , Cairong Zhao , Hanli Wang , Alexander Dvorkovich , Rui Fan This is my paper · ORCID
classification cs.CV
keywords unsupervisedstereomatchingdisparityconfidenceestimationrelativedepthpriorknowledgetransferrankinglosssmoothnessKITTIvisionfoundationmodels
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

The paper tries to show that unsupervised stereo matching can be made substantially more accurate by filtering disparity estimates with a confidence map derived from a monocular relative-depth prior, then transferring depth-ranking knowledge only through confident quasi-dense correspondences. It proposes a plug-and-play confidence estimator, DDCV, that votes on local coherence between neighboring disparities and their relative depths, plus two losses: a local depth ranking (LDR) loss and a dual disparity smoothness (DDS) loss. On the KITTI Stereo 2015 benchmark the resulting Un-ViTAStereo model reports D1-all of 5.03%, which the paper claims is state of the art among unsupervised stereo matching methods and a 22.7% relative improvement over the next-best unsupervised method. If true, the practical payoff is that an unsupervised stereo network can exploit the same frozen vision-foundation depth priors that have driven supervised methods.

What carries the argument

The central objects are three. First, DDCV: a parameter-free voting scheme that computes, for each pixel $p$, $C(p)=\frac{1}{m}\sum_{q\in\mathcal{N}_p} v_{pq}$ with $v_{pq}=F_{RC}(p,q)F_{VC}(p,q)$; $F_{RC}$ checks whether $(\tilde D(p)-\tilde D(q))(D(p)-D(q))$ is nonnegative, and $F_{VC}$ is defined to require mild disparity variations in stable-depth regions and large disparity variations at depth discontinuities. Second, the LDR loss: for each pixel, the top-$k$ highest-confidence disparities in a dilated window act as anchors, and the loss penalizes a target disparity only when its metric-depth ranking contradicts the relative-depth ranking of its anchor, with weights $\omega(\Delta\tilde D_{pr})$ favoring pairs at substantially different depths. Third, the DDS loss replaces image-gradient weighting in the standard smoothness loss with relative-depth-gradient weighting and adds a symmetric term that penalizes smooth disparities where the depth map jumps. The depth prior itself comes from a frozen vision Transformer (Depth Anything V2) that is shared, through ViTAStereo, with the stereo network.

What would settle it

Compute Eq. (3) on any real disparity and depth pair with the stated hyper-parameters: the first factor demands $|\Delta\tilde D_{pq}|\ge\gamma\sigma$ while the second demands $|\Delta\tilde D_{pq}|\le\gamma$, and with $\sigma>1$ no value satisfies both, so $C(p)=0$ for every pixel. Replacing DDCV's output by an all-zero confidence map while keeping LDR and DDS unchanged should reproduce the same EPE on SceneFlow if the confidence map contributes nothing.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is that the reliability of a disparity estimate can be judged by whether the local sign and magnitude changes in the disparity map agree with the local sign and magnitude changes in an off-the-shelf relative-depth map, and that using only high-confidence disparities as anchors for depth-ranking supervision removes the noise that plagues random-correspondence knowledge transfer. The authors package this into DDCV, whose confidence for a pixel is the average of binary votes from neighboring pixels; a positive vote requires both ranking consistency (Eq. 1) and variation consistency (Eq. 3), and Eq. 5 multiplies the two signals. The confidence map then selects top-k reference points for LDR, and DDS adds a depth-gradient-aware smoothness term that penalizes disagreement between disparity discontinuities and depth discontinuities. With these components, Un-ViTAStereo reports lower EPE and D1 than all published unsupervised methods on the KITTI Stereo benchmarks, and DDCV alone reaches AUC scores competitive with supervised confidence networks such as SEDNet.

Load-bearing premise

The method's confidence signal depends on Eq. (3), which as written asks a single vote to require both a large and a small depth change at the same time, so the vote is zero everywhere and the confidence map carries no information; if that formula is not corrected, the reported gains cannot come from DDCV.

Editorial extensions

If this is right

  • Any existing stereo network can adopt the framework without architecture or data changes, and at inference time only stereo image pairs are needed; the relative-depth map is required only during training.
  • Replacing random sparse correspondences with confidence-filtered quasi-dense correspondences should reduce the noise that previously degraded depth-ranking knowledge transfer; the SceneFlow ablation shows LDR alone lowers EPE by about 10.7% relative to the CDR loss.
  • The relative-depth-guided smoothness loss should improve disparity estimates at object boundaries in texture-less or repeated-pattern regions where image gradients mislead the standard edge-aware smoothness loss.
  • DDCV provides a label-free, real-time disparity confidence map that is competitive with supervised confidence estimators on KITTI, Middlebury and ETH3D across several disparity inputs.

Reading between the lines

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

  • A plausible correction of Eq. (3) that treats large-depth-change and stable-depth-change as two separate cases rather than simultaneous inequalities would make the variation-consistency vote nonzero; re-running the ablations with that corrected formula is the direct way to test whether the reported gains come from DDCV or from the two losses alone.
  • Because DDCV needs no labels and no training, it can be dropped into supervised stereo pipelines as a free confidence head; this would let future work decide whether its value in Table II, comparable AUC to SEDNet, translates into downstream error-detection gains.
  • The same rank-by-relative-depth, filter-by-confidence recipe could apply to other pixel-correspondence tasks, such as optical flow or video depth, wherever a monocular depth prior is available; the paper does not explore this.
  • If the KITTI results hold with a corrected confidence term, the practical divide between supervised and unsupervised stereo narrows, since the reported D1-all of 5.03% is within reach of some supervised figures on the same public benchmark; quantifying that gap is not attempted in the paper.
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 unsupervised stereo matching framework, Un-ViTAStereo, built on the authors' earlier ViTAStereo network. The three contributions are: (i) DDCV, a plug-and-play disparity confidence estimation algorithm that checks local coherence consistency between disparities and relative depths; (ii) a local depth ranking (LDR) loss that builds quasi-dense correspondences using only high-confidence disparities to transfer relative-depth ranking knowledge; and (iii) a dual disparity smoothness (DDS) loss that aligns disparity gradients with relative-depth gradients. Experiments on SceneFlow, KITTI 2012/2015, Middlebury, and ETH3D report state-of-the-art results among unsupervised stereo matching methods on the KITTI benchmarks, with extensive ablations and a runtime of 0.9 s for the full pipeline.

Significance. If the described mechanisms are correct, the work is a meaningful contribution to unsupervised stereo matching. The DDCV confidence estimator is novel in using relative depth priors, is parameter-free and real-time, and the LDR loss addresses a known weakness of prior knowledge-transfer methods by filtering noisy correspondences. The paper is also strong in scope: it provides public code and supplementary material, evaluates on multiple benchmarks, and includes careful ablations. However, the central formula for DDCV, Eq. (3), does not implement the behavior described in the text, and the reported performance gains are therefore not accounted for by the manuscript as written. This is a load-bearing issue that must be resolved before the claims can be accepted.

major comments (3)
  1. [III-A, Eq. (3)] The claim that F_VC is identically zero is not correct, because Θ(0)=1, so each factor in Eq. (3) evaluates to 1 whenever its inner condition is not met. However, the formula as written implements the opposite of the stated intention. For a pair with stable relative depth (|ΔeD| ≤ γ), the first factor is 1 but the second factor is 1 only if |ΔD| ≥ σ, i.e., it votes positively only for large disparity variations. For a pair at a depth discontinuity (|ΔeD| ≥ γσ), the second factor is 1 but the first factor is 1 only if |ΔD| ≤ 1, i.e., it votes positively only for small disparity variations. Thus DDCV assigns high confidence to disparity pairs that are inconsistent with the relative depth prior and low confidence to consistent pairs. Since the LDR loss in Eq. (10) selects the top-k highest-confidence disparities as reference points, the mechanism described cannot produce the behavior claimed in Sections I and III. The two Θ arguments in Eq. (3) appear to be interchanged relative to the description; please correct the formula and re-verify all experiments that depend on the DDCV confidence map.
  2. [IV-A, implementation details] The implementation states "We also set α = 2 in (3)", but Eq. (3) contains the symbol σ, not α. This is a transcription error that makes the reported hyper-parameter setting ambiguous. Please clarify whether σ = 2, and ensure that all symbols in the manuscript are consistent between the equations, the text, and the implementation details.
  3. [III-A and Fig. 2] The relative depth prior is generated by the Depth Anything V2 model, whose vision Transformer encoder is shared with the stereo matching network ViTAStereo, as shown in Fig. 2. The depth prior is therefore not fully independent of the features used for stereo matching; it is a frozen, pretrained representation, but the same encoder feeds both tasks. The paper should discuss the implications of this architectural sharing for the claimed "knowledge transfer" and the novelty of the approach. Ideally, an ablation with an alternative relative-depth source (e.g., a different monocular depth network) would demonstrate that the benefit is not an artifact of the shared representation.
minor comments (5)
  1. [II-A] Section II-A contains two consecutive, essentially identical paragraphs on unsupervised stereo matching; the duplication should be removed.
  2. [III-B1, Eq. (10)] The notation Θ(p, r) in Eq. (10) is undefined. Please define it explicitly (presumably as a step function on the consistency between ΔeD and ΔD, analogous to Eq. (1)).
  3. [III-B1, Eq. (10)] The denominator in Eq. (10) can be zero when all correspondences in the neighborhood are ranking-consistent. Please add a small epsilon or define the loss term as zero in that case to avoid division by zero.
  4. [Fig. 5] The axes in Fig. 5 are not labeled; the horizontal axis should be labeled with the number of reference points k and the vertical axis with EPE (pixel).
  5. [IV-D, Table II] Table II compares DDCV with supervised confidence estimation methods (CCNN, LGC, SEDNet), but DDCV additionally uses a relative depth map as input. This should be stated in the caption or text so that the comparison is not misread as head-to-head on identical inputs.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the relative-depth prior is externally trained, the proposed losses are evaluated against external KITTI benchmarks, and no fitted parameter is renamed as a prediction.

full rationale

The derivation chain is self-contained with respect to the claimed contribution. The relative depth prior is obtained from the externally trained Depth Anything V2 model [47]; the stereo backbone ViTAStereo is the authors' prior work, but it is used as a fixed architecture/baseline rather than as an unverified premise asserting that the new losses work. DDCV's confidence map is computed from the current disparity map and the external relative depth map via Eqs. (1)-(6); no parameter is fitted to ground-truth stereo disparities or to the reported KITTI errors. The LDR and DDS losses then train the network against the same external prior, and the central SoTA claim is evaluated against external KITTI Stereo benchmarks (Table IV) using published scores of other unsupervised methods. Self-citations to ViTAStereo and earlier stereo works are present but not load-bearing in the sense of importing a uniqueness theorem or pre-deciding the accuracy gain. Note that Eq. (3) appears internally unsatisfiable as written (the two Θ factors demand disjoint conditions on |ΔD| and |ΔeD| when σ>1), but that is a correctness/consistency flaw in the formulation, not a circularity; the evaluation itself does not reduce to the paper's own definitions.

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

The method introduces no new physical entities. Its central claims rest on several domain assumptions about the reliability of monocular depth priors and local coherence, plus five hand-chosen hyperparameters. The most fragile premise is the F_VC equation, which as written cannot produce a positive vote.

free parameters (5)
  • sigma = 2 (called alpha in Sect. IV-A)
    Variation consistency threshold in Eq. (3); chosen intuitively without extensive tuning. The paper inconsistently calls it alpha in the experiments.
  • lambda1, lambda2, lambda3 = 0.1 each
    Weights of the loss terms in Eq. (7); stated as intuitively chosen without extensive hyperparameter tuning.
  • k (top-k reference points) = 8
    Number of reference points in LDR; selected from the SceneFlow ablation where EPE first increases (Fig. 5).
  • sampling window size = 11x11
    Used in both DDCV and LDR; chosen intuitively in Sect. IV-A.
  • disparity stability threshold = 1 pixel
    Embedded in F_VC via (1 - |Delta D|); explicitly stated as an empirical assumption in the Conclusion.
assumptions (5)
  • domain assumption Relative depth maps from Depth Anything V2 provide reliable depth ordering between neighboring pixels.
    Used throughout Sect. III-A and III-B as the geometric prior that supervises disparity learning.
  • domain assumption Local coherence between neighboring disparities and relative depths indicates disparity correctness.
    Motivates DDCV and LDR in Sect. III-A; adopted from MRF-based stereo matching literature.
  • domain assumption Disparities change gradually across continuous regions.
    Explicitly invoked in Sect. III-B1 to justify the LDR loss design.
  • domain assumption Depth Anything V2 outputs larger values for closer objects, so disparity and relative depth are inversely proportional and this inversion is omitted.
    Stated in Sect. III-A to justify direct comparison of disparity and relative depth variations.
  • domain assumption Photometric and left-right consistency losses provide a valid baseline unsupervised training signal.
    These are the standard losses in unsupervised stereo matching, used as the baseline in Eq. (7) and Table I.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Integrating Disparity Confidence Estimation into Relative Depth Prior-Guided Unsupervised Stereo Matching." pith.science (2026). https://pith.science/paper/3HFKBB3L

@misc{pith2026250801275,
  author       = {Pith},
  title        = {Pith review of: Integrating Disparity Confidence Estimation into Relative Depth Prior-Guided Unsupervised Stereo Matching},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3HFKBB3L}},
  note         = {Machine review of arXiv:2508.01275}
}
read the original abstract

Unsupervised stereo matching has garnered significant attention for its independence from costly disparity annotations. Typical unsupervised methods rely on the multi-view consistency assumption for training networks, which suffer considerably from stereo matching ambiguities, such as repetitive patterns and texture-less regions. A feasible solution lies in transferring 3D geometric knowledge from a relative depth map to the stereo matching networks. However, existing knowledge transfer methods learn depth ranking information from randomly built sparse correspondences, which makes inefficient utilization of 3D geometric knowledge and introduces noise from mistaken disparity estimates. This work proposes a novel unsupervised learning framework to address these challenges, which comprises a plug-and-play disparity confidence estimation algorithm and two depth prior-guided loss functions. Specifically, the local coherence consistency between neighboring disparities and their corresponding relative depths is first checked to obtain disparity confidence. Afterwards, quasi-dense correspondences are built using only confident disparity estimates to facilitate efficient depth ranking learning. Finally, a dual disparity smoothness loss is proposed to boost stereo matching performance at disparity discontinuities. Experimental results demonstrate that our method achieves state-of-the-art stereo matching accuracy on the KITTI Stereo benchmarks among all unsupervised stereo matching methods.

Figures

Figures reproduced from arXiv: 2508.01275 by the authors.

Figure 1
Figure 1. (a) Left image, (b) disparity ground truth, and estimated disparity [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An illustration of the model architecture and training framework in this work. DPT represents the dense prediction Transformer [ [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Illustrations of inconsistency between image intensities and disparities [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Ablation study on the top-k operation in LLDR w.r.t. different k. as the Flying 3D test set, containing 4,370 stereo image pairs) with image resolution of 960 × 540 pixels. 2) Virtual KITTI [55] contains 21,260 stereo image pairs (resolution: 1, 242 × 375 pixels), gene…
Figure 6
Figure 6. Figure 6: Examples of disparity estimation results with or without our proposed [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Qualitative experimental results of CCNN [ [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Qualitative experimental results of Flow2Stereo [ [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

65 extracted references · 2 linked inside Pith

  1. [1]

    Playing to vision foundation model’s strengths in stereo matching,

    C.-W. Liu et al. , “Playing to vision foundation model’s strengths in stereo matching,” IEEE Transactions on Intelligent Vehicles , 2024, DOI:10.1109/TIV .2024.3467287

    arXiv 2024

  2. [2]

    SC-DepthV3: Robust self-supervised monocular depth estimation for dynamic scenes,

    L. Sun et al. , “SC-DepthV3: Robust self-supervised monocular depth estimation for dynamic scenes,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 46, no. 1, pp. 497–508, 2023

    2023

  3. [3]

    A resource-efficient pipelined architecture for real- time semi-global stereo matching,

    Z. Lu et al. , “A resource-efficient pipelined architecture for real- time semi-global stereo matching,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 2, pp. 660–673, 2021

    2021

  4. [4]

    Deep stereo network with mrf-based cost aggregation,

    K. Zeng et al., “Deep stereo network with mrf-based cost aggregation,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 34, no. 4, pp. 2426–2438, 2023

    2023

  5. [5]

    Stereo matching: fundamentals, state-of-the-art, and existing challenges,

    C.-W. Liu et al. , “Stereo matching: fundamentals, state-of-the-art, and existing challenges,” in Autonomous Driving Perception: Fundamentals and Applications. Springer, 2023, pp. 63–100

    2023

  6. [6]

    Road surface 3D reconstruction based on dense subpixel disparity map estimation,

    R. Fan et al., “Road surface 3D reconstruction based on dense subpixel disparity map estimation,” IEEE Transactions on Image Processing , vol. 27, no. 6, pp. 3025–3035, 2018

    2018

  7. [7]

    Learning collision-free space detection from stereo im- ages: Homography matrix brings better data augmentation,

    R. Fan et al. , “Learning collision-free space detection from stereo im- ages: Homography matrix brings better data augmentation,” IEEE/ASME Transactions on Mechatronics, vol. 27, no. 1, pp. 225–233, 2021

    2021

  8. [8]

    SNE-RoadSeg: Incorporating surface normal information into semantic segmentation for accurate freespace detection,

    R. Fan et al., “SNE-RoadSeg: Incorporating surface normal information into semantic segmentation for accurate freespace detection,” in Euro- pean Conference on Computer Vision (ECCV) . Springer, 2020, pp. 340–356

    2020

Show all 65 references
  1. [9]

    A low-cost and scalable framework to build large-scale localization benchmark for augmented reality,

    H. Liu et al. , “A low-cost and scalable framework to build large-scale localization benchmark for augmented reality,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 34, no. 4, pp. 2274– 2288, 2023

    2023

  2. [10]

    MASIC: Deep mask stereo image compression,

    X. Deng et al., “MASIC: Deep mask stereo image compression,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 33, no. 10, pp. 6026–6040, 2023

    2023

  3. [11]

    WHU-Stereo: A challenging benchmark for stereo matching of high-resolution satellite images,

    S. Li et al., “WHU-Stereo: A challenging benchmark for stereo matching of high-resolution satellite images,” IEEE Transactions on Geoscience and Remote Sensing , vol. 61, pp. 1–14, 2023

    2023

  4. [12]

    Pothole detection based on disparity transformation and road surface modeling,

    R. Fan et al. , “Pothole detection based on disparity transformation and road surface modeling,” IEEE Transactions on Image Processing , vol. 29, pp. 897–908, 2019

    2019

  5. [13]

    Graph attention layer evolves semantic segmentation for road pothole detection: A benchmark and algorithms,

    R. Fan et al., “Graph attention layer evolves semantic segmentation for road pothole detection: A benchmark and algorithms,” IEEE Transac- tions on Image Processing , vol. 30, pp. 8144–8154, 2021

    2021

  6. [14]

    A high-throughput depth estimation processor for accurate semiglobal stereo matching using pipelined inter-pixel aggregation,

    Y . Lee and H. Kim, “A high-throughput depth estimation processor for accurate semiglobal stereo matching using pipelined inter-pixel aggregation,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 1, pp. 411–422, 2021

    2021

  7. [15]

    Unambiguous pyramid cost volumes fusion for stereo matching,

    Q. Chen et al., “Unambiguous pyramid cost volumes fusion for stereo matching,” IEEE Transactions on Circuits and Systems for Video Tech- nology, vol. 34, no. 10, pp. 9223–9236, 2023

    2023

  8. [16]

    Rethinking road surface 3-D reconstruction and pothole detection: From perspective transformation to disparity map segmenta- tion,

    R. Fan et al., “Rethinking road surface 3-D reconstruction and pothole detection: From perspective transformation to disparity map segmenta- tion,” IEEE Transactions on Cybernetics, vol. 52, no. 7, pp. 5799–5808, 2022

    2022

  9. [17]

    The farther the better: Balanced stereo matching via depth-based sampling and adaptive feature refinement,

    H. Zhang et al. , “The farther the better: Balanced stereo matching via depth-based sampling and adaptive feature refinement,” IEEE Transac- tions on Circuits and Systems for Video Technology , vol. 32, no. 7, pp. 4613–4625, 2021

    2021

  10. [18]

    Inter-scale similarity guided cost aggregation for stereo matching,

    P. Li et al. , “Inter-scale similarity guided cost aggregation for stereo matching,” IEEE Transactions on Circuits and Systems for Video Tech- nology, vol. 35, no. 1, pp. 134–147, 2024

    2024

  11. [19]

    Robust scale-aware stereo matching network,

    J. Okae et al. , “Robust scale-aware stereo matching network,” IEEE Transactions on Artificial Intelligence, vol. 3, no. 2, pp. 244–253, 2021

    2021

  12. [20]

    A decomposition model for stereo matching,

    C. Yao et al. , “A decomposition model for stereo matching,” in Pro- ceedings of the Computer Vision and Pattern Recognition Conference (CVPR), 2021, pp. 6091–6100

    2021

  13. [21]

    Stereo anywhere: Robust zero-shot deep stereo matching even where either stereo or mono fail,

    L. Bartolomei et al. , “Stereo anywhere: Robust zero-shot deep stereo matching even where either stereo or mono fail,” in Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR) , 2025, pp. 1013–1027

    2025

  14. [22]

    Monster: Marry monodepth to stereo unleashes power,

    J. Cheng et al., “Monster: Marry monodepth to stereo unleashes power,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025, pp. 6273–6282

    2025

  15. [23]

    Unsupervised stereo matching using confidential correspondence consistency,

    S. Joung et al. , “Unsupervised stereo matching using confidential correspondence consistency,” IEEE Transactions on Intelligent Trans- portation Systems, vol. 21, no. 5, pp. 2190–2203, 2019

    2019

  16. [24]

    SegStereo: Exploiting semantic information for disparity estimation,

    G. Yang et al., “SegStereo: Exploiting semantic information for disparity estimation,” in Proceedings of the European Conference on Computer Vision (ECCV), 2018, pp. 636–651

    2018

  17. [25]

    SUW-learn: Joint supervised, unsupervised, weakly supervised deep learning for monocular depth estimation,

    H. Ren et al. , “SUW-learn: Joint supervised, unsupervised, weakly supervised deep learning for monocular depth estimation,” in Proceed- 12 IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition...

    2020

  18. [26]

    Unsupervised deep event stereo for depth estima- tion,

    S. N. Uddin et al. , “Unsupervised deep event stereo for depth estima- tion,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 11, pp. 7489–7504, 2022

    2022

  19. [27]

    Self-supervised learning of PSMNet via generative adversarial networks,

    X. Yang et al. , “Self-supervised learning of PSMNet via generative adversarial networks,” in International Conference on Intelligent Com- puting (ICIC). Springer, 2024, pp. 469–479

    2024

  20. [28]

    Self-supervised learning for stereo matching with self-improving ability,

    Y . Zhong et al. , “Self-supervised learning for stereo matching with self-improving ability,” Computing Research Repository (CoRR) , vol. abs/1709.00930, 2017. [Online]. Available: https://arxiv.org/abs/1709. 00930

    2017 arXiv

  21. [29]

    Occlusion aware stereo matching via coopera- tive unsupervised learning,

    A. Li and Z. Yuan, “Occlusion aware stereo matching via coopera- tive unsupervised learning,” in Asian Conference on Computer Vision . Springer, 2018, pp. 197–213

    2018

  22. [30]

    Unsupervised occlusion-aware stereo matching with directed disparity smoothing,

    A. Li et al. , “Unsupervised occlusion-aware stereo matching with directed disparity smoothing,” IEEE Transactions on Intelligent Trans- portation Systems, vol. 23, no. 7, pp. 7457–7468, 2021

    2021

  23. [31]

    Co-Teaching: An ark to unsupervised stereo matching,

    H. Wang et al., “Co-Teaching: An ark to unsupervised stereo matching,” in 2021 IEEE International Conference on Image Processing (ICIP) . IEEE, 2021, pp. 3328–3332

    2021

  24. [32]

    Flow2Stereo: Effective self-supervised learning of optical flow and stereo matching,

    P. Liu et al., “Flow2Stereo: Effective self-supervised learning of optical flow and stereo matching,” in Proceedings of the IEEE/CVF conference on Computer Vision and Pattern Recognition (CVPR) , 2020, pp. 6648– 6657

    2020

  25. [33]

    On the confidence of stereo matching in a deep-learning era: a quantitative evaluation,

    M. Poggi et al., “On the confidence of stereo matching in a deep-learning era: a quantitative evaluation,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 44, no. 9, pp. 5293–5313, 2021

    2021

  26. [34]

    A quantitative evaluation of confidence measures for stereo vision,

    X. Hu and P. Mordohai, “A quantitative evaluation of confidence measures for stereo vision,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 34, no. 11, pp. 2121–2133, 2012

    2012

  27. [35]

    Quantitative evaluation of confidence measures in a machine learning world,

    M. Poggi et al. , “Quantitative evaluation of confidence measures in a machine learning world,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV) , 2017, pp. 5228–5237

    2017

  28. [36]

    Learning a general-purpose confidence measure based on O(1) features and a smarter aggregation strategy for semi global matching,

    M. Poggi and S. Mattoccia, “Learning a general-purpose confidence measure based on O(1) features and a smarter aggregation strategy for semi global matching,” in 2016 Fourth International Conference on 3D Vision (3DV). IEEE, 2016, pp. 509–518

    2016

  29. [37]

    Learning a confidence measure in the disparity domain from O(1) features,

    M. Poggi et al., “Learning a confidence measure in the disparity domain from O(1) features,” Computer Vision and Image Understanding , vol. 193, pp. 2905–2913, 2020

    2020

  30. [38]

    Self-adapting confidence estimation for stereo,

    M. Poggi et al. , “Self-adapting confidence estimation for stereo,” in Proceedings of the European Conference on Computer Vision (ECCV) . Springer, 2020, pp. 715–733

    2020

  31. [39]

    Learning from scratch a confidence measure,

    M. Poggi and S. Mattoccia, “Learning from scratch a confidence measure,” in Proceedings of 27th British Machine Vision Conference (BMVC), 2016, pp. 1–13

    2016

  32. [40]

    Beyond local reasoning for stereo confidence estimation with deep learning,

    F. Tosi et al., “Beyond local reasoning for stereo confidence estimation with deep learning,” in Proceedings of the European Conference on Computer Vision (ECCV) , 2018, pp. 319–334

    2018

  33. [41]

    U-Net: Convolutional networks for biomedical image segmentation,

    O. Ronneberger et al., “U-Net: Convolutional networks for biomedical image segmentation,” in Proceddings of the Medical Image Computing and Cmputer-Assisted Intervention (MICCAI). Springer, 2015, pp. 234– 241

    2015

  34. [42]

    Learning the distribution of errors in stereo matching for joint disparity and uncertainty estimation,

    L. Chen et al. , “Learning the distribution of errors in stereo matching for joint disparity and uncertainty estimation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 17 235–17 244

    2023

  35. [43]

    Learning confidence measures in the wild,

    F. Tosi et al., “Learning confidence measures in the wild,” inProceedings of 28th British Machine Vision Conference (BMVC) , 2017, pp. 1–13

    2017

  36. [44]

    Learning and selecting confidence measures for robust stereo matching,

    M.-G. Park and K.-J. Yoon, “Learning and selecting confidence measures for robust stereo matching,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 41, no. 6, pp. 1397–1411, 2018

    2018

  37. [45]

    Using self-contradiction to learn confidence mea- sures in stereo vision,

    C. Mostegel et al. , “Using self-contradiction to learn confidence mea- sures in stereo vision,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2016, pp. 4067– 4076

    2016

  38. [46]

    Vision Transformer adapter for dense predictions,

    Z. Chen et al. , “Vision Transformer adapter for dense predictions,” in International Conference on Learning Representations (ICLR) , 2023

    2023

  39. [47]

    Depth Anything V2,

    L. Yang et al., “Depth Anything V2,” Advances in Neural Information Processing Systems (NeurIPS) , vol. 37, pp. 21 875–21 911, 2024

    2024

  40. [48]

    Segment anything,

    A. Kirillov et al., “Segment anything,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , 2023, pp. 4015– 4026

    2023

  41. [49]

    Depth Anything: Unleashing the power of large- scale unlabeled data,

    L. Yang et al. , “Depth Anything: Unleashing the power of large- scale unlabeled data,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2024, pp. 10 371– 10 381

    2024

  42. [50]

    Stereo without epipolar lines: A maximum-flow formulation,

    S. Roy, “Stereo without epipolar lines: A maximum-flow formulation,” International Journal of Computer Vision , vol. 34, no. 2, pp. 147–161, 1999

    1999

  43. [51]

    These maps are made by propagation: Adapting deep stereo networks to road scenarios with decisive disparity diffusion,

    C.-W. Liu et al., “These maps are made by propagation: Adapting deep stereo networks to road scenarios with decisive disparity diffusion,” IEEE Transactions on Image Processing, vol. 34, pp. 1516–1528, 2025

    2025

  44. [52]

    PVStereo: Pyramid voting module for end-to-end self- supervised stereo matching,

    H. Wang et al., “PVStereo: Pyramid voting module for end-to-end self- supervised stereo matching,” IEEE Robotics and Automation Letters , vol. 6, no. 3, pp. 4353–4360, 2021

    2021

  45. [53]

    Image quality assessment: from error visibility to structural similarity,

    Z. Wang et al. , “Image quality assessment: from error visibility to structural similarity,” IEEE Transactions on Image Processing , vol. 13, no. 4, pp. 600–612, 2004

    2004

  46. [54]

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

    N. Mayer et al. , “A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2016, pp. 4040–4048

    2016

  47. [55]

    Virtual KITTI 2,

    Y . Cabon et al. , “Virtual KITTI 2,” Computing Research Repository (CoRR), vol. abs/2001.10773, 2020. [Online]. Available: https: //arxiv.org/abs/2001.10773

    2001 arXiv

  48. [56]

    Are we ready for autonomous driving? the KITTI vision benchmark suite,

    A. Geiger et al. , “Are we ready for autonomous driving? the KITTI vision benchmark suite,” in 2012 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . IEEE, 2012, pp. 3354–3361

    2012

  49. [57]

    Object scene flow for autonomous vehicles,

    M. Menze and A. Geiger, “Object scene flow for autonomous vehicles,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp. 3061–3070

    2015

  50. [58]

    High-resolution stereo datasets with subpixel- accurate ground truth,

    D. Scharstein et al. , “High-resolution stereo datasets with subpixel- accurate ground truth,” in Pattern Recognition: 36th German Conference (GCPR). Springer, 2014, pp. 31–42

    2014

  51. [59]

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

    T. Schops et al., “A multi-view stereo benchmark with high-resolution images and multi-camera videos,” in Proceedings of the IEEE Confer- ence on Computer Vision and Pattern Recognition (CVPR) , 2017, pp. 3260–3269

    2017

  52. [60]

    Iterative geometry encoding volume for stereo matching,

    G. Xu et al., “Iterative geometry encoding volume for stereo matching,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023, pp. 21 919–21 928

    2023

  53. [61]

    Practical stereo matching via cascaded recurrent network with adaptive correlation,

    J. Li et al. , “Practical stereo matching via cascaded recurrent network with adaptive correlation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 16 263– 16 272

    2022

  54. [62]

    A permutation model for the self- supervised stereo matching problem,

    P.-A. Brousseau and S. Roy, “A permutation model for the self- supervised stereo matching problem,” in 2022 19th Conference on Robots and Vision (CRV) . IEEE, 2022, pp. 122–131

    2022

  55. [63]

    Parallax attention for unsupervised stereo correspon- dence learning,

    L. Wang et al. , “Parallax attention for unsupervised stereo correspon- dence learning,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 4, pp. 2108–2125, 2020

    2020

  56. [64]

    Adaptive cost volume representation for unsuper- vised high-resolution stereo matching,

    K. W. Tong et al. , “Adaptive cost volume representation for unsuper- vised high-resolution stereo matching,” IEEE Transactions on Intelligent Vehicles, vol. 8, no. 1, pp. 912–922, 2022

    2022

  57. [65]

    A novel cell structure-based disparity estimation for unsupervised stereo matching,

    X. Cheng et al. , “A novel cell structure-based disparity estimation for unsupervised stereo matching,” IET Image Processing , vol. 16, no. 6, pp. 1678–1693, 2022. Chuang-Wei Liu received his B.E. degree in au- tomation from Tongji University in 2020. He is currently pursuing ...

    2022

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.