Pith. sign in

REVIEW 4 major objections 5 minor 31 references

Improving Optical Flow and Stereo Depth Estimation by Leveraging Uncertainty-Based Learning Difficulties

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims that weighting each pixel's training loss by its difficulty—more weight for hard-to-match pixels, less for occluded ones—improves optical flow and stereo depth accuracy on standard benchmarks.

desk verdict A plausible but single-run training recipe whose modest gains are not yet separated from noise; worth a careful review but not citable as established. read the letter →

arxiv 2506.00324 v1 pith:AJFMWHRA submitted 2025-05-31 cs.CV

Jisoo Jeong , Hong Cai , Jamie Menjay Lin , Fatih Porikli This is my paper
classification cs.CV
keywords opticalflowstereodepthestimationconfidencemapuncertaintyocclusioncycleconsistencytraininglossRAFT
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

Most optical flow and stereo depth networks are trained with the same loss applied to every pixel, even though some pixels are far harder to learn than others. This paper proposes two training losses that exploit that variation: a Difficulty Balancing (DB) loss, which gives extra weight to pixels where the network's current prediction is far from the ground truth, and an Occlusion Avoiding (OA) loss, which gives less weight to regions that fail a forward-backward consistency check and are therefore likely occluded. Combining the two losses, multiplicatively for optical flow and with a masked sum for stereo, is reported to improve accuracy over the RAFT, FlowFormer, and RAFT-Stereo baselines on Sintel, KITTI, ETH3D, and Middlebury. A reader should care because the change is confined to the loss, leaving the network architecture untouched, so the recipe could transfer to any matching model.

What carries the argument

The central objects are two confidence maps. $M_{DB}$ is computed as $\exp(-\|f_{gt}-f\|_2^2)$ from the predicted and ground-truth flow (Eq. 3), marking pixels the model already fits well with values near 1. $M_{OA}$ is computed from forward-backward consistency (Eq. 6), marking regions where matching is reliable and likely non-occluded. The DB loss weights each pixel's L1 error by $1+\alpha(1-M_{DB})^{\beta}$, so hard, low-confidence pixels get more gradient; the OA loss weights by $1+\alpha M_{OA}^{\beta}$, so occluded, inconsistent pixels stay near their standard weight while confident pixels get more. For stereo, the reverse right-to-left disparity needed for $M_{OA}$ is obtained by flipping the stereo pair, estimating disparity, and flipping back, so the same network supplies both directions. Combining the two maps multiplicatively (optical flow) or with a hard mask plus sum (stereo) prevents the two opposite weighting tendencies from canceling in occluded regions.

What would settle it

Retrain each baseline and each proposed loss at least five times from different random seeds and compare average end-point error with confidence intervals; if the intervals overlap on most benchmarks, the claimed improvements are not established.

Watch

Extended reading notes

Core claim

The paper's central claim is that pixel-wise learning difficulty in dense correspondence tasks has two separable causes, and that a loss treating them separately outperforms a uniform loss. The first cause is low model confidence: pixels whose predicted flow or disparity is far from the ground truth are harder, so the DB loss multiplies their L1 error by a factor $1+\alpha(1-M_{DB})^{\beta}$, where $M_{DB}=\exp(-\|f_{gt}-f\|_2^2)$ is an error-based confidence map. The second cause is occlusion: pixels without a real correspondence should not be forced to match, so the OA loss uses a cycle-consistency map $M_{OA}$ from Eq. (6) and weights the error by $1+\alpha M_{OA}^{\beta}$, keeping occluded regions near the standard L1 weight while boosting confident regions. The paper reports that the multiplicative combination is best for optical flow, the mask-sum combination is best for stereo, and that the combined loss improves over all three baselines.

Load-bearing premise

The reported gains come from single training runs with no error bars, so the paper assumes that differences of about 0.01 to 0.1 pixels in average error (for example, 1.43 vs 1.35 on Sintel clean) are real improvements rather than run-to-run training noise.

Editorial extensions

If this is right

  • Applied to RAFT, the combined loss reduces Sintel clean EPE from 1.43 to 1.35 and KITTI Fl-all from 17.4 to 15.45, according to the paper's Table 1.
  • Applied to FlowFormer, the multiplicative combination lowers Sintel clean EPE from 1.01 to 0.97 and KITTI Fl-all from 14.72 to 14.17.
  • Applied to RAFT-Stereo, the mask-sum combination lowers ETH3D bad-1.0 from 2.44 to 1.67 and KITTI D1-all from 1.96 to 1.83 on the reported test sets.
  • The DB loss mainly helps large-displacement and object-boundary pixels, while OA mainly helps small-displacement and occluded regions, as separated in Table 5.

Reading between the lines

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

  • Because the losses change only the training objective, they could be dropped into other dense matching networks, such as scene flow or video depth models, where hard pixels and occlusions also vary spatially; the paper does not test this.
  • The same cycle-consistency map used for OA could serve as a self-supervised weight in settings without ground truth, making the recipe applicable to semi-supervised or unsupervised training; this is an extension, not a claim of the paper.
  • The hyperparameters $\alpha$ and $\beta$ are chosen by validation grid search; an annealed schedule over training, where the focus shifts from hard pixels to occlusion avoidance, is a natural variation the authors do not explore.
  • Since the benchmarks are single runs, a practical adopter would first replicate the comparison over multiple seeds before deciding; the paper itself reports no variance statistics.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes two training losses for optical flow and stereo depth estimation: a Difficulty Balancing (DB) loss that reweights the per-pixel L1 loss by an error-based confidence map, and an Occlusion Avoiding (OA) loss that reweights by a forward-backward cycle-consistency confidence map. Several combinations of the two losses are tested, and the authors select multiplicative combination for optical flow and mask-sum for stereo. The losses are applied to RAFT, FlowFormer, and RAFT-Stereo and evaluated on Sintel, KITTI, ETH3D, and Middlebury, with reported improvements over the baselines on most metrics.

Significance. If the reported improvements are reproducible, the contribution is practically useful: it provides simple, model-agnostic training losses that can be plugged into existing recurrent optical flow and stereo architectures. The paper is thorough in its ablation coverage, including hyperparameter grids for both losses and four combination rules across two tasks and several benchmarks. The main weakness is that the central evidence consists of single training runs with differences of 0.01-0.1 EPE on Sintel, which is within the range of typical run-to-run variance for these models. The significance of the claimed improvement is therefore conditional on additional variance analysis and a clear statement of which configuration produced each reported number.

major comments (4)
  1. [Section 4.2, Tables 1 and 5] The headline optical-flow gains are small (e.g., 1.43 to 1.35 EPE on Sintel clean, 2.71 to 2.65 on Sintel final, 5.04 to 4.50 on KITTI EPE), yet every configuration is reported from a single training run with no error bars or multiple seeds. For recurrent models trained with stochastic optimization, these deltas are within the range of run-to-run variation, so the Abstract and Section 6 claim of 'significant performance improvements' is not supported by the reported evidence. Please provide multiple seeds per configuration, or a paired statistical test, to establish that the gains exceed training noise. The same concern applies to the stereo results in Table 3.
  2. [Sections 3.1 and 5.4] The DB hyperparameter specification is internally inconsistent. Section 3.1 sets (alpha, beta) = (2.0, 0.5) for optical flow and points to Tables 6 and 8; Table 1's DB row exactly matches the (2.0, 0.5) entry in Table 6 (1.41 / 2.68 / 4.65 / 15.92). However, Section 5.4 states that '(alpha = 2.0, beta = 1.0) shows the best overall accuracy' and that this configuration improves over the RFL loss. The reader cannot determine which configuration produced the headline DB and combination results. Please correct the inconsistency and state the exact loss configuration used for every reported result.
  3. [Section 5.4, Tables 6-8] The hyperparameters alpha and beta, as well as the combination rule (multiplication for optical flow, mask-sum for stereo), are selected by evaluating on the same Sintel/KITTI/ETH3D/Middlebury splits that are then used to claim improvement in Tables 1-4. Because the reported numbers are selected maxima over a grid, they are optimistically biased and cannot serve as independent evidence of improvement. Please either fix the configuration before evaluation on a held-out split or report the full grid and clearly describe the selection procedure in the text.
  4. [Table 2] The finetuned model on Sintel test shows an overall EPE-all regression on the clean pass (1.609 to 1.685), with unmatched EPE worsening from 9.647 to 10.367. The text emphasizes matched-pixel improvements and describes 'significant improvement in the matching area,' but it does not explain or contextualize the overall clean-pass regression. This is directly relevant to the claim of consistent improvement and should be addressed in the discussion.
minor comments (5)
  1. [Section 4.1] The setup paragraph contains a typo: 'Florformer' should be 'FlowFormer'.
  2. [Section 2.1] In the description of RAFT-stereo, 'achieves' is misspelled as 'achieves'; please correct the typo.
  3. [Table 3 header] The dataset name is misspelled as 'MiddlueBury'; it should be 'Middlebury'.
  4. [Section 5.4] The OA ablation text says 'We apply five different αs ... to RAFT-Stereo model,' but the table reports optical flow results on RAFT; this should say 'RAFT model.' Also, 'TableTable 8' should be 'Table 8.'
  5. [Section 3.2] The OA loss for stereo requires an additional reverse-disparity pass (the transformation-and-restoration technique in Figure 3), but the training-time computational overhead of this extra pass is not quantified. Please state the additional cost so practitioners can evaluate the trade-off.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: the losses are training-time reweightings and the claimed improvements are external benchmark measurements; only minor self-citation and benchmark-selection concerns are present.

full rationale

No circular derivation chain was found. The DB and OA losses are training-time reweighting schemes applied to standard optical-flow and stereo losses, and the reported accuracies are measurements on external benchmarks (Sintel, KITTI, ETH3D, Middlebury) compared against published baselines such as RAFT, FlowFormer, and RAFT-Stereo. The claimed improvements are therefore not equivalent by construction to the loss definitions. The error-based confidence map (Eq. 3) uses ground truth at training time, which is ordinary supervised weighting, not a prediction derived from itself. The cycle-consistency confidence map (Eq. 6) is inherited from DistractFlow and ultimately from the standard forward-backward consistency check of UnFlow, and it is applied as an input weighting rather than as the quantity being predicted. The paper does select hyperparameters and combination forms on the Sintel/KITTI train sets in Tables 6-8 and then reports those same sets as evidence of improvement, which is a benchmark-selection concern, but it is not circularity: the headline numbers are measurements, not quantities defined by the fitted parameters. Self-citations (DistractFlow, OCAI, SCIFlow) are disclosed and serve as sources of the confidence-map machinery; the central empirical claim does not reduce to those citations. An internal inconsistency exists between Section 3.1, which specifies (alpha=2.0, beta=0.5) for optical-flow DB, and Section 5.4, which states that (alpha=2.0, beta=1.0) shows the best overall accuracy, while the DB row of Table 1 matches the (2.0, 0.5) entry of Table 6. That is a reproducibility and correctness flaw, not a circularity. The score of 1 reflects minor self-citation and the selection-on-evaluation-set concern; neither makes the derivation self-referential.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on two confidence maps: an error-based confidence map (Eq. 3, borrowed from LiteFlowNetV3) and a cycle-consistency confidence map (Eq. 6, borrowed from DistractFlow and UnFlow). The paper's own contribution is the weighting formulas and the combination rule, with hyperparameters (alpha,beta) and combination choice all selected on benchmark validation sets. No new entities are introduced; all entries are training-time weights.

free parameters (5)
  • DB loss alpha, beta (optical flow) = alpha=2.0, beta=0.5
    Selected by grid search on Sintel (train) and KITTI (train), Table 6.
  • DB loss alpha, beta (stereo) = alpha=2.0, beta=1.0
    Selected by grid search on ETH3D, Middlebury, KITTI, Table 8.
  • OA loss alpha, beta (optical flow) = alpha=2.0, beta=1.0
    Selected by grid search, Table 7.
  • OA loss alpha, beta (stereo) = alpha=1.0, beta=1.0
    Selected by grid search, Table 8.
  • Combination rule = multiplication for OF, mask-sum for stereo
    Chosen by comparing four combination variants on validation metrics, Sections 3.3 and 5.3.
assumptions (4)
  • domain assumption Cycle consistency based confidence map (Eq. 6) reliably indicates occlusion and matchable regions during training.
    Used in OA loss (Eq. 8); taken from DistractFlow [7] and UnFlow [22] without validation on current tasks.
  • domain assumption Error based confidence (Eq. 3) computed from ground truth is a useful proxy for per-pixel learning difficulty.
    Underlies DB loss (Eq. 7).
  • domain assumption The transformation-and-restoration (flip and swap) of stereo images yields an accurate right-to-left disparity map.
    Used in Section 3.2 and Fig 3 to produce MOA for stereo; assumes no significant artifacts from flipping.
  • domain assumption Hyperparameters tuned on Sintel (train) and KITTI (train) generalize to the respective test sets and to other benchmarks.
    All hyperparameter choices are validated on these sets; test set improvements are reported for a subset of configurations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving Optical Flow and Stereo Depth Estimation by Leveraging Uncertainty-Based Learning Difficulties." pith.science (2026). https://pith.science/paper/AJFMWHRA

@misc{pith2026250600324,
  author       = {Pith},
  title        = {Pith review of: Improving Optical Flow and Stereo Depth Estimation by Leveraging Uncertainty-Based Learning Difficulties},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AJFMWHRA}},
  note         = {Machine review of arXiv:2506.00324}
}
read the original abstract

Conventional training for optical flow and stereo depth models typically employs a uniform loss function across all pixels. However, this one-size-fits-all approach often overlooks the significant variations in learning difficulty among individual pixels and contextual regions. This paper investigates the uncertainty-based confidence maps which capture these spatially varying learning difficulties and introduces tailored solutions to address them. We first present the Difficulty Balancing (DB) loss, which utilizes an error-based confidence measure to encourage the network to focus more on challenging pixels and regions. Moreover, we identify that some difficult pixels and regions are affected by occlusions, resulting from the inherently ill-posed matching problem in the absence of real correspondences. To address this, we propose the Occlusion Avoiding (OA) loss, designed to guide the network into cycle consistency-based confident regions, where feature matching is more reliable. By combining the DB and OA losses, we effectively manage various types of challenging pixels and regions during training. Experiments on both optical flow and stereo depth tasks consistently demonstrate significant performance improvements when applying our proposed combination of the DB and OA losses.

Figures

Figures reproduced from arXiv: 2506.00324 by the authors.

Figure 1
Figure 1. (a) Most existing methods (e.g., [19, 28]) predominately treat training loss on each pixel equally for optical flow and Stereo depth. (b) SCIFlow [18] utilizes a Regression Focal Loss, which focuses more on low-confident samples, for training optical flow models. (c) Our proposed approach more comprehensively con￾siders two sources of learning difficulties in training, i.e., model confidence and occlusion. When trai… view at source ↗
Figure 2
Figure 2. Overview of our method. Optical Flows (f1→2 and f2→1) or Disparity (dL→R and dR→L) are computed by the same model for the consecutive or stereo image pair. Error map based Confidence map is obtained using prediction and ground truth (Eq. 3). Forward backward consistency based Confidence map is computed by Eq. 6. These confidence maps are used in the training loss. ∗ represents the combination of two losses. where γ1… view at source ↗
Figure 3
Figure 3. The Transformation-and-Restoration technique to ob￾tain the reverse (right-to-left) disparity. sight to guide the network regarding occlusions: \label {oa_loss} \footnotesize \begin {split} OF \quad \ : \quad l_{OA}^{i} = || (1 + \alpha \cdot (M_{OA})^{\beta }) \cdot (f_{gt} - f_{1\rightarrow 2}^{i})||_{1}, \\ Stereo : \quad l_{OA}^{i} = || (1 + \alpha \cdot (M_{OA})^{\beta }) \cdot (d_{gt} - d_{L \rightarrow R}^{i}… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Optical flow qualitative results on KITTI (train) using RAFT and our models. First row is for the baseline. Second and third rows [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Stereo Depth Qualitative results on KITTI (train) using RAFT-Stereo and our models. First row is the Baseline results. Second [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Confidence map results of MDB and MOA. Error based confidence map MDB (left) is obtained by Eq. 3, and 1−MDB (middle) is used in the loss function. Forward backward consistency based confidence map MOA (right) is computed by 6. a 0.1 Dl-all difference on the KITTI (tra…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

31 extracted references · 28 canonical work pages

  1. [1]

    A naturalistic open source movie for optical flow evaluation

    Daniel J Butler, Jonas Wulff, Garrett B Stanley, and Michael J Black. A naturalistic open source movie for optical flow evaluation. In Proceedings of the European Conference on Computer Vision, pages 611–625. Springer, 2012. 4

    work page 2012

  2. [2]

    Temporal hockey action recognition via pose and optical flows

    Zixi Cai, Helmut Neher, Kanav Vats, David A Clausi, and John Zelek. Temporal hockey action recognition via pose and optical flows. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition Work- shops, pages 0–0, 2019. 1

    work page 2019

  3. [3]

    Flownet: Learning optical flow with convolutional networks

    Alexey Dosovitskiy, Philipp Fischer, Eddy Ilg, Philip Hausser, Caner Hazirbas, Vladimir Golkov, Patrick Van Der Smagt, Daniel Cremers, and Thomas Brox. Flownet: Learning optical flow with convolutional networks. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 2758–2766, 2015. 4

    work page 2015

  4. [4]

    Vision meets robotics: The kitti dataset

    Andreas Geiger, Philip Lenz, Christoph Stiller, and Raquel Urtasun. Vision meets robotics: The kitti dataset. The Inter- national Journal of Robotics Research , 32(11):1231–1237,

  5. [5]

    Flowformer: A transformer architecture for optical flow

    Zhaoyang Huang, Xiaoyu Shi, Chao Zhang, Qiang Wang, Ka Chun Cheung, Hongwei Qin, Jifeng Dai, and Hongsheng Li. Flowformer: A transformer architecture for optical flow. In Proceedings of the European Conference on Computer Vi- sion, 2022. 1, 2, 4, 5

    work page 2022

  6. [6]

    Liteflownet3: Resolv- ing correspondence ambiguity for more accurate optical flow estimation

    Tak-Wai Hui and Chen Change Loy. Liteflownet3: Resolv- ing correspondence ambiguity for more accurate optical flow estimation. In Computer Vision–ECCV 2020: 16th Euro- pean Conference, Glasgow, UK, August 23–28, 2020, Pro- ceedings, Part XX 16, pages 169–184. Springer, 2020. 2

    work page 2020

  7. [7]

    Distractflow: Improving optical flow estimation via real- istic distractions and pseudo-labeling

    Jisoo Jeong, Hong Cai, Risheek Garrepalli, and Fatih Porikli. Distractflow: Improving optical flow estimation via real- istic distractions and pseudo-labeling. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13691–13700, 2023. 2, 3

    work page 2023

  8. [8]

    Ocai: Improving optical flow estimation by occlusion and consistency aware interpolation

    Jisoo Jeong, Hong Cai, Risheek Garrepalli, Jamie Menjay Lin, Munawar Hayat, and Fatih Porikli. Ocai: Improving optical flow estimation by occlusion and consistency aware interpolation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19352– 19362, 2024. 1, 3

    work page 2024

Show all 31 references
  1. [9]

    Learning to estimate hidden motions with global motion aggregation

    Shihao Jiang, Dylan Campbell, Yao Lu, Hongdong Li, and Richard Hartley. Learning to estimate hidden motions with global motion aggregation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 9772– 9781, 2021. 1, 2

    2021

  2. [10]

    What matters in unsupervised optical flow

    Rico Jonschkowski, Austin Stone, Jonathan T Barron, Ariel Gordon, Kurt Konolige, and Anelia Angelova. What matters in unsupervised optical flow. InProceedings of the European Conference on Computer Vision , pages 557–572. Springer,

  3. [11]

    Moving object tracking using optical flow and motion vector estima- tion

    Kiran Kale, Sushant Pawar, and Pravin Dhulekar. Moving object tracking using optical flow and motion vector estima- tion. In 2015 4th international conference on reliability, in- focom technologies and optimization (ICRITO)(trends and future directions), pages 1–6. IEEE, 2015. 1

    2015

  4. [12]

    Dy- namicstereo: Consistent dynamic depth from stereo videos

    Nikita Karaev, Ignacio Rocco, Benjamin Graham, Natalia Neverova, Andrea Vedaldi, and Christian Rupprecht. Dy- namicstereo: Consistent dynamic depth from stereo videos. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13229–13239, 2023. 1

    2023

  5. [13]

    The hci benchmark suite: Stereo and flow ground truth with uncertainties for urban autonomous driv- ing

    Daniel Kondermann, Rahul Nair, Katrin Honauer, Karsten Krispin, Jonas Andrulis, Alexander Brock, Burkhard Gusse- feld, Mohsen Rahimimoghaddam, Sabine Hofmann, Claus Brenner, et al. The hci benchmark suite: Stereo and flow ground truth with uncertainties for urban autonomous dr...

    2016

  6. [14]

    Ifrnet: Intermediate feature refine network for efficient frame interpolation

    Lingtong Kong, Boyuan Jiang, Donghao Luo, Wenqing Chu, Xiaoming Huang, Ying Tai, Chengjie Wang, and Jie Yang. Ifrnet: Intermediate feature refine network for efficient frame interpolation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , p...

    1969

  7. [15]

    Motion feature network: Fixed motion filter for action recognition

    Myunggi Lee, Seungeui Lee, Sungjoon Son, Gyutae Park, and Nojun Kwak. Motion feature network: Fixed motion filter for action recognition. In Proceedings of the European Conference on Computer Vision, pages 387–403, 2018. 1

    2018

  8. [16]

    Practical stereo matching via cascaded re- current network with adaptive correlation

    Jiankun Li, Peisen Wang, Pengfei Xiong, Tao Cai, Zi- wei Yan, Lei Yang, Jiangyu Liu, Haoqiang Fan, and Shuaicheng Liu. Practical stereo matching via cascaded re- current network with adaptive correlation. In Proceedings of the IEEE/CVF conference on computer vision and pattern...

    2022

  9. [17]

    Amt: All-pairs multi-field transforms for efficient frame interpolation

    Zhen Li, Zuo-Liang Zhu, Ling-Hao Han, Qibin Hou, Chun- Le Guo, and Ming-Ming Cheng. Amt: All-pairs multi-field transforms for efficient frame interpolation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 9801–9810, 2023. 1

    2023

  10. [18]

    Sciflow: Empowering lightweight optical flow models with self-cleaning iterations

    Jamie Menjay Lin, Jisoo Jeong, Hong Cai, Risheek Gar- repalli, Kai Wang, and Fatih Porikli. Sciflow: Empowering lightweight optical flow models with self-cleaning iterations. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2162–2171,...

    2024

  11. [19]

    Raft-stereo: Multilevel recurrent field transforms for stereo matching

    Lahav Lipson, Zachary Teed, and Jia Deng. Raft-stereo: Multilevel recurrent field transforms for stereo matching. In 2021 International Conference on 3D Vision (3DV) , pages 218–227. IEEE, 2021. 1, 2, 4

    2021

  12. [20]

    Dvc: An end-to-end deep video com- pression framework

    Guo Lu, Wanli Ouyang, Dong Xu, Xiaoyun Zhang, Chunlei Cai, and Zhiyong Gao. Dvc: An end-to-end deep video com- pression framework. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 11006–11015, 2019. 1

    2019

  13. [21]

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

    Nikolaus Mayer, Eddy Ilg, Philip Hausser, Philipp Fischer, Daniel Cremers, Alexey Dosovitskiy, and Thomas Brox. A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation. In Proceedings of the IEEE/CVF Conference on Computer Vision ...

    2016

  14. [22]

    Unflow: Un- supervised learning of optical flow with a bidirectional cen- 9 sus loss

    Simon Meister, Junhwa Hur, and Stefan Roth. Unflow: Un- supervised learning of optical flow with a bidirectional cen- 9 sus loss. In Proceedings of the AAAI conference on artificial intelligence, 2018. 2

    2018

  15. [23]

    Object scene flow for autonomous vehicles

    Moritz Menze and Andreas Geiger. Object scene flow for autonomous vehicles. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 3061–3070, 2015. 4

    2015

  16. [24]

    High-resolution stereo datasets with subpixel-accurate ground truth

    Daniel Scharstein, Heiko Hirschm ¨uller, York Kitajima, Greg Krathwohl, Nera Ne ˇsi´c, Xi Wang, and Porter West- ling. High-resolution stereo datasets with subpixel-accurate ground truth. In Pattern Recognition: 36th German Confer- ence, GCPR 2014, M¨unster, Germany, September...

    2014

  17. [25]

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

    Thomas Schops, Johannes L Schonberger, Silvano Galliani, Torsten Sattler, Konrad Schindler, Marc Pollefeys, and An- dreas Geiger. A multi-view stereo benchmark with high- resolution images and multi-camera videos. In Proceed- ings of the IEEE conference on computer vision and ...

    2017

  18. [26]

    Fixmatch: Simplifying semi-supervised learning with consistency and confidence

    Kihyuk Sohn, David Berthelot, Nicholas Carlini, Zizhao Zhang, Han Zhang, Colin A Raffel, Ekin Dogus Cubuk, Alexey Kurakin, and Chun-Liang Li. Fixmatch: Simplifying semi-supervised learning with consistency and confidence. Advances in neural information processing systems, 33:5...

    2020

  19. [27]

    Smurf: Self-teaching multi- frame unsupervised raft with full-image warping

    Austin Stone, Daniel Maurer, Alper Ayvaci, Anelia An- gelova, and Rico Jonschkowski. Smurf: Self-teaching multi- frame unsupervised raft with full-image warping. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3887–3896, 2021. 2

    2021

  20. [28]

    Raft: Recurrent all-pairs field transforms for optical flow

    Zachary Teed and Jia Deng. Raft: Recurrent all-pairs field transforms for optical flow. In Proceedings of the European Conference on Computer Vision , pages 402–419. Springer,

  21. [29]

    Video compression through image interpolation

    Chao-Yuan Wu, Nayan Singhal, and Philipp Krahenbuhl. Video compression through image interpolation. InProceed- ings of the European Conference on Computer Vision, pages 416–431, 2018. 1

    2018

  22. [30]

    Separable flow: Learning motion cost volumes for optical flow estimation

    Feihu Zhang, Oliver J Woodford, Victor Adrian Prisacariu, and Philip HS Torr. Separable flow: Learning motion cost volumes for optical flow estimation. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 10807–10817, 2021. 2

    2021

  23. [31]

    Maskflownet: Asymmetric feature matching with learnable occlusion mask

    Shengyu Zhao, Yilun Sheng, Yue Dong, Eric I Chang, Yan Xu, et al. Maskflownet: Asymmetric feature matching with learnable occlusion mask. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 6278–6287, 2020. 1, 3 10

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.