Pith. sign in

REVIEW 3 major objections 5 minor 74 references

CodingHomo: Bootstrapping Deep Homography With Video Coding

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read Motion vectors already embedded in H.264 video, normally discarded at decode time, can be fused with image features through learned masks to give unsupervised homography its best reported accuracy: 0.31 on CA-unsup and 1.21 on GHOF.

desk verdict Solid niche advance using H.264 MVs for unsupervised homography; credible but needs a clearer GHOF training protocol and validation-based alpha. read the letter →

arxiv 2504.12165 v1 pith:EKVBD4PE submitted 2025-04-16 cs.CV

Yike Liu , Haipeng Li , Shuaicheng Liu , Bing Zeng This is my paper · ORCID
classification cs.CV
keywords homographyestimationunsupervisedlearningmotionvectorsH.264/AVCvideocodingpriormask-guidedfusiondominantplaneimagealignment
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

Motion vectors computed by an H.264 encoder to compress video are usually thrown away after decoding; this paper argues that they are a free, reliable prior for estimating the homography between consecutive frames. CodingHomo is an unsupervised network that compares these blockwise motion vectors with the current homography flow, learns masks that mark which motion agrees with the dominant plane, and uses the masked information both to fuse motion into the estimator and to focus the training loss. If the argument is right, video-enabled devices get improved image alignment at no extra bitrate and without labels, exactly in the conditions where appearance-based methods struggle: low light, low texture, fog, rain, snow, and moving foreground objects. On CA-unsup the method reports the best average point-matching error among the compared learning-based methods, 0.31 versus 0.34 for the previous best, and on the unseen GHOF test set it reports 1.21 versus 1.72.

What carries the argument

The load-bearing mechanism is a pair of learned masks built by comparing two motion fields: the blockwise motion-vector flow $V^i_{ab}$ extracted from the H.264 bitstream and the current homography flow $H^i_{ab}$. In the Mask-Guided Fusion (MGF) module, a mask generator $G_m$ outputs a motion-rejection mask $M^i_m$ whose confidence values are modeled with conditionally independent Laplace distributions, and a fusion network predicts a residual homography added to $H^i_{ab}$. The Mask-Guided Homography Estimation (MGHE) module applies an adjusted mask $\hat{M}^{i-1}_m = M^{i-1}_m(1-\alpha)+\alpha$ to the warped source and target features before a Swin transformer refinement, so early noisy estimates cannot zero out the features. Finally, the Enhanced Motion Mask $M_e = G_c(F'_a,F_b)\cdot G_m(H_{ab},V_{ab})$ multiplies a coplanarity-aware feature mask with the motion-rejection mask and weights the unsupervised loss, trained with a negative log-likelihood term that keeps the mask from collapsing to all zeros.

What would settle it

Replace the H.264 motion-vector input with a scrambled or all-zero flow on the CA-unsup test set and compare average point-matching error: if it remains near 0.31 the motion prior is not load-bearing, and if it jumps toward the 0.36-0.46 range seen in the no-mask ablations, the prior is confirmed as the source of the gain.

Watch

Extended reading notes

Core claim

The paper's central claim is that motion vectors already present in coded video, which are ordinarily discarded after decoding, constitute an effective and entirely free prior for unsupervised homography estimation, and that a network can learn to use them by masking. CodingHomo predicts a homography from reconstructed H.264 frame pairs plus their motion-vector flows; the mask-guided fusion and mask-guided estimation modules suppress foreground and parallax motion while the enhanced motion mask concentrates the loss on the dominant plane. The paper reports that this reaches an average point-matching error of 0.31 on CA-unsup, a reduction from 0.34 for the previous best method, and 1.21 on GHOF, a reduction from 1.72, including generalization to fog, rain, and snow scenes not present in training.

Load-bearing premise

The masks only help if the codec's motion vectors and the current homography agree about the dominant plane; when the initial homography is far wrong or the true motion is larger than the codec's short search window, the prior stops being informative and the method has no backup signal.

Editorial extensions

If this is right

  • Because the motion vectors are extracted from the bitstream during normal decoding, the method adds no bitrate cost and no new labeling; any pipeline that already decodes H.264 video can obtain the prior for free.
  • The reported generalization to fog, rain, and snow on GHOF, categories absent from training, indicates that the MV prior transfers across appearance domains better than photometric or feature losses alone.
  • The ablation numbers attribute most of the gain to the learned masks: removing the motion-rejection mask raises CA-unsup error from 0.31 to 0.36, and removing the mask from the loss raises it to 0.46, so the mask mechanism, not the raw MVs, is what carries the improvement.
  • The coarse-to-fine estimator inherits a transformer pyramid, but the MGHE module lets the mask prune features before each refinement stage, which is what the paper credits for handling large foregrounds and multiple planes.
  • The approach is stated to extend naturally to HEVC/H.265 and VVC/H.266, since MV extraction is a standard decoding step; the H.264 choice is made because the CA-unsup dataset was encoded in that format.

Reading between the lines

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

  • Editorial extension: the same mask-guided fusion pattern could be applied to other decode-side priors, such as residual energy, block partition structure, or reference-frame indexes, to distinguish static background from dynamic objects; the paper does not explore these.
  • Editorial boundary: because H.264 motion search is typically capped around a 16-pixel window, the method is inherently a small-baseline estimator, and feature-based methods may remain necessary for large displacements unless codec search ranges are enlarged.
  • Editorial test: swapping the H.264 MVs for optical flow computed from the decoded frames would show whether the gain comes specifically from the codec prior or from any dense motion signal; the paper's mechanism does not logically require the MVs to originate from a codec.
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 presents CodingHomo, an unsupervised deep homography estimation framework that augments image pairs with motion vectors extracted from H.264/AVC video bitstreams. The main technical contributions are a Mask-Guided Fusion (MGF) module that fuses a coarse homography with motion vectors under a learned motion-rejection mask, a Mask-Guided Homography Estimation (MGHE) module that uses the mask in coarse-to-fine refinement, and an Enhanced Motion Mask (EMM) that combines a coplanarity-aware mask with a motion-rejection mask for loss weighting, trained with a negative log-likelihood loss. Experiments on the CA-unsup dataset report state-of-the-art average point matching error (0.31 PME) and a generalization experiment on the GHOF dataset reports a larger improvement (1.21 PME), with qualitative results on fog, rain, and snow scenes. Code and dataset are promised publicly.

Significance. If the empirical results are reproducible, this is a novel and useful direction: it is the first unsupervised homography framework to exploit codec motion vectors as a prior, and the engineering is solid in several respects. The paper includes comparisons against feature-based, supervised, and unsupervised baselines on two datasets, per-component ablations, standard deviations, a t-test, robustness curves, and a public code/data commitment. These strengths make the central claim clearly testable. The main risks are protocol-level: the GHOF generalization experiment does not state whether GHOF training frames were used in any way, and the hyperparameter α is selected on the test set. Both are fixable with clarifications or additional experiments, so the paper is best treated as a major revision rather than a rejection.

major comments (3)
  1. [Section IV-C2 / Section IV-A] The GHOF generalization experiment lacks an explicit training protocol. The paper reports GHOF as a generalization test and emphasizes improvements on unseen fog, rain, and snow scenes, but Section IV-A states that GHOF includes a training set of approximately 10,000 frames, and Section IV-C2 only says "we added the corresponding MVs to the test set (GHOF)". It is never stated whether the network, the mask generators, or the hyperparameter α were trained or tuned on any GHOF frames. Since the largest reported relative gain (1.72 to 1.21, 29.65%) occurs on the GHOF categories, this omission leaves open the possibility that the improvement reflects learning GHOF-specific MV/homography statistics rather than a generalizable motion prior. Please state definitively which GHOF data were used for training, validation, and model selection; if none were used, say so explicitly.
  2. [Section IV-F / Table IV] The floor parameter α in Eq. (6) is selected on the CA-unsup test set. Table IV reports PME on the test set for α ∈ {1.0, 0.4, 0.2, 0.0}, and the final model uses α = 0.2 from that table. This is test-set model selection, so the PME values in Table I and the t-test against RealSH (t = 3.7102) are obtained from a model whose hyperparameter was tuned on the same test set, while the comparison methods were not afforded such tuning. Please introduce a validation split, select α on it, and report the corresponding test numbers, or report all α variants on a held-out validation set to establish that the SOTA claim is not an artifact of test-set tuning.
  3. [Section IV-C1] The statement "All learning-based methods are trained on the CA-sup dateset" is inconsistent with the rest of the manuscript, where CAHomo, BasesHomo, and HomoGAN are described as unsupervised methods trained on CA-unsup. If some comparison methods were trained on a different dataset (for example RealSH on CA-sup), the comparison in Tables I and II is not controlled; if the sentence is a typo, it should be corrected. Please specify the exact training data and protocol for every learning-based baseline, including RealSH and the unsupervised methods.
minor comments (5)
  1. [Section III-B] Please fix the typo "A VC/H.264" (should be "AVC/H.264") and the repeated "video video coding" in the same section.
  2. [Table I] The row labeled "I3×3" is not defined in the text or caption; please state explicitly that it is the identity-transform baseline.
  3. [Figure 2 caption] The caption says "Red block donates a dynamic vehicle"; "donates" should be "denotes".
  4. [Section IV-A / IV-C2] The encoder configuration used to generate the H.264 bitstreams and MVs for GHOF is not reported. The paper only says MVs were "added" to GHOF; without encoder settings (QP, search range, frame structure, minimum block size, reference-frame count), the motion-vector input is not reproducible.
  5. [Section III-D] The mask notation is inconsistent: "Mi−1m", "Mi m", and "fMi−1m" are used without a uniform convention. Please define the superscript/subscript placement clearly and use it consistently through Eqs. (5)–(7).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claim is an empirical benchmark result evaluated on held-out ground truth, and the MVs and masks are internal inputs/training objectives rather than predictions that reproduce their own inputs.

full rationale

The paper's central claim is an empirical benchmark result: CodingHomo is trained with an unsupervised photometric/NLL objective on CA-unsup training pairs and evaluated by average L2 point-matching error against held-out ground-truth correspondences on the CA-unsup and GHOF test sets. The motion vectors are codec inputs extracted from the bitstream, not labels derived from the target homography; the learned masks are internal gating modules whose supervision (Eqs. 4, 8, 12) depends on the current estimate H_ab and the codec flow V_ab, but this is a training objective, not a prediction of the test metric. No equation defines the reported PME in terms of the training loss, and no fitted parameter is renamed as a prediction. The GHOF generalization experiment states that fog/rain scenes 'were not included in the training dataset (CA-unsup)' and that MVs were added only to the test set; whether that protocol was fully followed is a reproducibility question, not a circularity one. Self-citations to CAHomo, HomoGAN, RealSH, and GHOF are used as baselines, datasets, and borrowed backbone components; they are not invoked as uniqueness theorems or as the justification for the empirical improvement. The claimed derivation chain is therefore self-contained with respect to circularity.

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

The central claim rests on the availability and informativeness of codec MVs. The most consequential free parameter is alpha, tuned on the test set via ablation. The main domain assumptions are about H.264 metadata availability and the small-baseline motion regime.

free parameters (3)
  • alpha (mask floor in MGHE) = 0.2
    Chosen via ablation on the CA-unsup test set (Table IV). Performance ranges from 0.31 (alpha=0.2) to 0.38 (alpha=0.0) and 0.34 (alpha=1.0); the claim of superiority depends on this choice.
  • BCE loss weight in plane loss = 0.05
    Fixed in Eq (14) without ablation; controls the all-zero-mask avoidance and affects mask quality.
  • Number of coarse-to-fine scale levels = 3
    Architectural choice reported in Implementation Details; no ablation shown.
assumptions (4)
  • domain assumption H.264 motion vectors can be extracted losslessly during decoding and represent inter-frame block motion.
    Sec III-B and Fig 4: the entire input signal V_ab depends on the codec metadata being available; for raw or differently coded video the method cannot be applied as described.
  • ad hoc to paper Motion vector flow and homography flow are commensurate representations, so their pixel-wise difference can be modeled by a Laplace distribution whose variance is the mask.
    Eq (4) in Sec III-C defines the likelihood used in the NLL training loss; this probabilistic link is a modeling choice, not derived from the codec or geometry.
  • domain assumption The dominant-plane homography is the transformation that aligns the largest image region, and MVs in that region are consistent with the homography.
    Assumed throughout Sec III-C to III-E; if the scene is dominated by a non-planar or non-rigid structure, the mask may reject the true alignment target.
  • domain assumption The MVs in the test videos have a bounded search range (around 16 pixels), so the method is only valid for small-baseline motions.
    Stated as a limitation in Sec V; for large baselines, MVs are uninformative and the method cannot improve over baselines.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CodingHomo: Bootstrapping Deep Homography With Video Coding." pith.science (2026). https://pith.science/paper/EKVBD4PE

@misc{pith2026250412165,
  author       = {Pith},
  title        = {Pith review of: CodingHomo: Bootstrapping Deep Homography With Video Coding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EKVBD4PE}},
  note         = {Machine review of arXiv:2504.12165}
}
read the original abstract

Homography estimation is a fundamental task in computer vision with applications in diverse fields. Recent advances in deep learning have improved homography estimation, particularly with unsupervised learning approaches, offering increased robustness and generalizability. However, accurately predicting homography, especially in complex motions, remains a challenge. In response, this work introduces a novel method leveraging video coding, particularly by harnessing inherent motion vectors (MVs) present in videos. We present CodingHomo, an unsupervised framework for homography estimation. Our framework features a Mask-Guided Fusion (MGF) module that identifies and utilizes beneficial features among the MVs, thereby enhancing the accuracy of homography prediction. Additionally, the Mask-Guided Homography Estimation (MGHE) module is presented for eliminating undesired features in the coarse-to-fine homography refinement process. CodingHomo outperforms existing state-of-the-art unsupervised methods, delivering good robustness and generalizability. The code and dataset are available at: \href{github}{https://github.com/liuyike422/CodingHomo

Figures

Figures reproduced from arXiv: 2504.12165 by the authors.

Figure 1
Figure 1. The overview of our work. We extract MVs while decoding frames [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An example of network input. (a) Reconstruct image pair. (b) MVs. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The overall pipeline of CodingHomo. Our network architecture consists of three stages: 1) Feature extraction stage. A CNN module for projecting [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Illustration of the extraction of MVs during the decoding process. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: The mask-guided fusion (MGF) module’s structure involves utilizing [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 7
Figure 7. Figure 7: The mask-guided homography estimation (MGHE) module. Adjust f [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: Qualitative results of our method and other existing learning-based methods on the CA-unsup [ [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 9
Figure 9. Figure 9: Qualitative results of feature-based methods and our method. For each feature-based method, we show its results with the best performed outlier [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Qualitative results of our method and other competitive methods on the GHOF [ [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]
Figure 11
Figure 11. Figure 11: Visualization results of outlier rejection masks of our method and other competitive methods on the CA-unsup [ [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]
Figure 12
Figure 12. Figure 12: The proportion of inliers of our method and all comparison methods under various thresholds. Inliers indicate points with errors under the threshold. [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]
Figure 13
Figure 13. Figure 13: Illustration of masks. The yellow box indicates that an interrupted area is finally eliminated in the fused masks. [PITH_FULL_IMAGE:figures/full_fig_p012_13.png]
Figure 14
Figure 14. Figure 14: Illustration of the masks generated without/with NLL constraint. [PITH_FULL_IMAGE:figures/full_fig_p013_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

74 extracted references · 59 canonical work pages

  1. [1]

    Hartley and A

    R. Hartley and A. Zisserman, Multiple view geometry in computer vision. Cambridge university press, 2003

    2003

  2. [2]

    Homography loss for monocular 3d object detection,

    J. Gu, B. Wu, L. Fan, J. Huang, S. Cao, Z. Xiang, and X.-S. Hua, “Homography loss for monocular 3d object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 1080–1089

    work page 2022

  3. [3]

    Homography guided temporal fusion for road line and marking segmentation,

    S. Wang, C. Nguyen, J. Liu, K. Zhang, W. Luo, Y . Zhang, S. Muthu, F. A. Maken, and H. Li, “Homography guided temporal fusion for road line and marking segmentation,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 1075–1085

    work page 2023

  4. [4]

    Learning collision-free space detection from stereo images: Homogra- phy matrix brings better data augmentation,

    R. Fan, H. Wang, P. Cai, J. Wu, M. J. Bocus, L. Qiao, and M. Liu, “Learning collision-free space detection from stereo images: Homogra- phy matrix brings better data augmentation,” IEEE/ASME Transactions on Mechatronics, vol. 27, no. 1, pp. 225–233, 2021

    work page 2021

  5. [5]

    Homography- based robust pose compensation and fusion imaging for augmented reality based endoscopic navigation system,

    W. Li, J. Fan, S. Li, Z. Tian, D. Ai, H. Song, and J. Yang, “Homography- based robust pose compensation and fusion imaging for augmented reality based endoscopic navigation system,” Computers in Biology and Medicine, vol. 138, p. 104864, 2021

    work page 2021

  6. [6]

    Homography-based visual servoing with remote center of motion for semi-autonomous robotic endoscope manipulation,

    M. Huber, J. B. Mitchell, R. Henry, S. Ourselin, T. Vercauteren, and C. Bergeles, “Homography-based visual servoing with remote center of motion for semi-autonomous robotic endoscope manipulation,” in 2021 International Symposium on Medical Robotics (ISMR) . IEEE, 2021, pp. 1–7

    work page 2021

  7. [7]

    Stereo correspon- dence and reconstruction of endoscopic data challenge,

    M. Allan, J. Mcleod, C. Wang, J. C. Rosenthal, Z. Hu, N. Gard, P. Eisert, K. X. Fu, T. Zeffiro, W. Xia et al. , “Stereo correspon- dence and reconstruction of endoscopic data challenge,” arXiv preprint arXiv:2101.01133, 2021

    arXiv 2021

  8. [8]

    Deep high dynamic range imaging with large foreground motions,

    S. Wu, J. Xu, Y .-W. Tai, and C.-K. Tang, “Deep high dynamic range imaging with large foreground motions,” in Proceedings of the European Conference on Computer Vision (ECCV) , 2018, pp. 117–132. 14

    work page 2018

Show all 74 references
  1. [9]

    Adnet: Attention-guided deformable convolutional network for high dynamic range imaging,

    Z. Liu, W. Lin, X. Li, Q. Rao, T. Jiang, M. Han, H. Fan, J. Sun, and S. Liu, “Adnet: Attention-guided deformable convolutional network for high dynamic range imaging,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 463– 470

    2021

  2. [10]

    Ghost-free high dynamic range imaging with context-aware transformer,

    Z. Liu, Y . Wang, B. Zeng, and S. Liu, “Ghost-free high dynamic range imaging with context-aware transformer,” in Proceedings of the European Conference on Computer Vision (ECCV) . Springer, 2022, pp. 344–360

    2022

  3. [11]

    Fast burst images denoising,

    Z. Liu, L. Yuan, X. Tang, M. Uyttendaele, and J. Sun, “Fast burst images denoising,” ACM Transactions on Graphics (TOG) , vol. 33, no. 6, pp. 1–9, 2014

    2014

  4. [12]

    Deep burst super- resolution,

    G. Bhat, M. Danelljan, L. Van Gool, and R. Timofte, “Deep burst super- resolution,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021, pp. 9209–9218

    2021

  5. [13]

    Bundled camera paths for video stabilization,

    S. Liu, L. Yuan, P. Tan, and J. Sun, “Bundled camera paths for video stabilization,” ACM Transactions on Graphics (TOG), vol. 32, no. 4, pp. 1–10, 2013

    2013

  6. [14]

    Minimum latency deep online video stabilization,

    Z. Zhang, Z. Liu, P. Tan, B. Zeng, and S. Liu, “Minimum latency deep online video stabilization,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 23 030–23 039

    2023

  7. [15]

    Gl-rg: Global-local representation granularity for video captioning,

    L. Yan, Q. Wang, Y . Cui, F. Feng, X. Quan, X. Zhang, and D. Liu, “Gl-rg: Global-local representation granularity for video captioning,” in Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, IJCAI-22 , 2022, pp. 2769–2775

    2022

  8. [16]

    Video captioning using global-local representation,

    L. Yan, S. Ma, Q. Wang, Y . Chen, X. Zhang, A. Savakis, and D. Liu, “Video captioning using global-local representation,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 10, pp. 6642– 6656, 2022

    2022

  9. [17]

    Solve the puzzle of instance segmentation in videos: A weakly supervised framework with spatio-temporal collaboration,

    L. Yan, Q. Wang, S. Ma, J. Wang, and C. Yu, “Solve the puzzle of instance segmentation in videos: A weakly supervised framework with spatio-temporal collaboration,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 1, pp. 393–406, 2022

    2022

  10. [18]

    Sg-net: Spatial granularity network for one-stage video instance segmentation,

    D. Liu, Y . Cui, W. Tan, and Y . Chen, “Sg-net: Spatial granularity network for one-stage video instance segmentation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 9816–9825

    2021

  11. [19]

    Tf-blender: Temporal feature blender for video object detection,

    Y . Cui, L. Yan, Z. Cao, and D. Liu, “Tf-blender: Temporal feature blender for video object detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 8138–8147

    2021

  12. [20]

    Video object detection for autonomous driving: Motion-aid feature calibration,

    D. Liu, Y . Cui, Y . Chen, J. Zhang, and B. Fan, “Video object detection for autonomous driving: Motion-aid feature calibration,” Neurocomput- ing, vol. 409, pp. 1–11, 2020

    2020

  13. [21]

    Distinctive image features from scale-invariant keypoints,

    D. G. Lowe, “Distinctive image features from scale-invariant keypoints,” International Journal of Computer Vision , vol. 60, pp. 91–110, 2004

    2004

  14. [22]

    Speeded-up robust features (surf),

    H. Bay, A. Ess, T. Tuytelaars, and L. Van Gool, “Speeded-up robust features (surf),” Computer Vision and Image Understanding , vol. 110, no. 3, pp. 346–359, 2008

    2008

  15. [23]

    Orb: An efficient alternative to sift or surf,

    E. Rublee, V . Rabaud, K. Konolige, and G. Bradski, “Orb: An efficient alternative to sift or surf,” in Proceedings of the IEEE/CVF International Conference on Computer Vision . Ieee, 2011, pp. 2564–2571

    2011

  16. [24]

    Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,

    M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Communications of the ACM, vol. 24, no. 6, pp. 381–395, 1981

    1981

  17. [25]

    Content-aware unsupervised deep homography estimation,

    J. Zhang, C. Wang, S. Liu, L. Jia, N. Ye, J. Wang, J. Zhou, and J. Sun, “Content-aware unsupervised deep homography estimation,” in Proceedings of the European Conference on Computer Vision (ECCV) . Springer, 2020, pp. 653–669

    2020

  18. [26]

    Gyroflow+: Gyroscope-guided unsupervised deep homography and optical flow learning,

    H. Li, K. Luo, B. Zeng, and S. Liu, “Gyroflow+: Gyroscope-guided unsupervised deep homography and optical flow learning,” International Journal of Computer Vision , pp. 1–19, 2024

    2024

  19. [27]

    Deep image homogra- phy estimation,

    D. DeTone, T. Malisiewicz, and A. Rabinovich, “Deep image homogra- phy estimation,” arXiv preprint arXiv:1606.03798 , 2016

    2016 arXiv

  20. [28]

    Localtrans: A multiscale local transformer network for cross-resolution homography estimation,

    R. Shao, G. Wu, Y . Zhou, Y . Fu, L. Fang, and Y . Liu, “Localtrans: A multiscale local transformer network for cross-resolution homography estimation,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 14 890–14 899

    2021

  21. [29]

    Iterative deep homography estimation,

    S.-Y . Cao, J. Hu, Z. Sheng, and H.-L. Shen, “Iterative deep homography estimation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 1879–1888

    2022

  22. [30]

    Supervised homography learning with realistic dataset generation,

    H. Jiang, H. Li, S. Han, H. Fan, B. Zeng, and S. Liu, “Supervised homography learning with realistic dataset generation,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 9806–9815

    2023

  23. [31]

    Microsoft coco: Common objects in context,

    T.-Y . Lin, M. Maire, S. Belongie, J. Hays, P. Perona, D. Ramanan, P. Doll ´ar, and C. L. Zitnick, “Microsoft coco: Common objects in context,” in Proceedings of the European Conference on Computer Vision (ECCV). Springer, 2014, pp. 740–755

    2014

  24. [32]

    Unsupervised deep homography: A fast and robust homography esti- mation model,

    T. Nguyen, S. W. Chen, S. S. Shivakumar, C. J. Taylor, and V . Kumar, “Unsupervised deep homography: A fast and robust homography esti- mation model,” IEEE Robotics and Automation Letters , vol. 3, no. 3, pp. 2346–2353, 2018

    2018

  25. [33]

    Motion basis learning for unsupervised deep homography estimation with subspace projection,

    N. Ye, C. Wang, H. Fan, and S. Liu, “Motion basis learning for unsupervised deep homography estimation with subspace projection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 13 117–13 125

    2021

  26. [34]

    Unsupervised homography estimation with coplanarity-aware gan,

    M. Hong, Y . Lu, N. Ye, C. Lin, Q. Zhao, and S. Liu, “Unsupervised homography estimation with coplanarity-aware gan,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 17 663–17 672

    2022

  27. [35]

    Codingflow: Enable video coding for video stabilization,

    S. Liu, M. Li, S. Zhu, and B. Zeng, “Codingflow: Enable video coding for video stabilization,” IEEE Transactions on Image Processing, vol. 26, no. 7, pp. 3291–3302, 2017

    2017

  28. [36]

    Overview of the h. 264/avc video coding standard,

    T. Wiegand, G. J. Sullivan, G. Bjontegaard, and A. Luthra, “Overview of the h. 264/avc video coding standard,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 13, no. 7, pp. 560–576, 2003

    2003

  29. [37]

    Overview of the high efficiency video coding (hevc) standard,

    G. J. Sullivan, J.-R. Ohm, W.-J. Han, and T. Wiegand, “Overview of the high efficiency video coding (hevc) standard,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 22, no. 12, pp. 1649– 1668, 2012

    2012

  30. [38]

    Overview of the versatile video coding (vvc) standard and its applications,

    B. Bross, Y .-K. Wang, Y . Ye, S. Liu, J. Chen, G. J. Sullivan, and J.- R. Ohm, “Overview of the versatile video coding (vvc) standard and its applications,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 31, no. 10, pp. 3736–3764, 2021

    2021

  31. [39]

    Discrete cosine transform,

    N. Ahmed, T. Natarajan, and K. R. Rao, “Discrete cosine transform,” IEEE Transactions on Computers , vol. 100, no. 1, pp. 90–93, 1974

    1974

  32. [40]

    Context-based adaptive binary arithmetic coding in the h. 264/avc video compression standard,

    D. Marpe, H. Schwarz, and T. Wiegand, “Context-based adaptive binary arithmetic coding in the h. 264/avc video compression standard,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 13, no. 7, pp. 620–636, 2003

    2003

  33. [41]

    Adaptive deblocking filter,

    P. List, A. Joch, J. Lainema, G. Bjontegaard, and M. Karczewicz, “Adaptive deblocking filter,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 13, no. 7, pp. 614–619, 2003

    2003

  34. [42]

    Dvc: An end-to-end deep video compression framework,

    G. Lu, W. Ouyang, D. Xu, X. Zhang, C. Cai, and Z. Gao, “Dvc: An end-to-end deep video compression framework,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2019, pp. 11 006–11 015

    2019

  35. [43]

    M-lvc: Multiple frames prediction for learned video compression,

    J. Lin, D. Liu, H. Li, and F. Wu, “M-lvc: Multiple frames prediction for learned video compression,” in Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, 2020, pp. 3546–3554

    2020

  36. [44]

    Fvc: A new framework towards deep video compression in feature space,

    Z. Hu, G. Lu, and D. Xu, “Fvc: A new framework towards deep video compression in feature space,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021, pp. 1502–1511

    2021

  37. [45]

    Displacement measurement and its application in interframe image coding,

    J. Jain and A. Jain, “Displacement measurement and its application in interframe image coding,” IEEE Transactions on Communications , vol. 29, no. 12, pp. 1799–1808, 1981

    1981

  38. [46]

    Interframe coding that follows the motion,

    Y . Taki, M. Hatori, and S. Tanaka, “Interframe coding that follows the motion,” Proc. Institute of Electronics and Communication Engineers Jpn. Annu. Conv.(IECEJ), p. 1263, 1974

    1974

  39. [47]

    Motion-compensated interframe coding for video conferenc- ing,

    T. Koga, “Motion-compensated interframe coding for video conferenc- ing,” in Proc. National Telecommunications Conference, 1981, 1981, pp. G5–3

    1981

  40. [48]

    A new three-step search algorithm for block motion estimation,

    R. Li, B. Zeng, and M. L. Liou, “A new three-step search algorithm for block motion estimation,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 4, no. 4, pp. 438–442, 1994

    1994

  41. [49]

    A novel four-step search algorithm for fast block motion estimation,

    L.-M. Po and W.-C. Ma, “A novel four-step search algorithm for fast block motion estimation,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 6, no. 3, pp. 313–317, 1996

    1996

  42. [50]

    A new diamond search algorithm for fast block- matching motion estimation,

    S. Zhu and K.-K. Ma, “A new diamond search algorithm for fast block- matching motion estimation,” IEEE Transactions on Image Processing , vol. 9, no. 2, pp. 287–290, 2000

    2000

  43. [51]

    K-nearest neighbour classifiers-a tutorial,

    P. Cunningham and S. J. Delany, “K-nearest neighbour classifiers-a tutorial,” ACM Computing Surveys (CSUR) , vol. 54, no. 6, pp. 1–25, 2021

    2021

  44. [52]

    Lift: Learned invariant feature transform,

    K. M. Yi, E. Trulls, V . Lepetit, and P. Fua, “Lift: Learned invariant feature transform,” in Proceedings of the European Conference on Computer Vision (ECCV) . Springer, 2016, pp. 467–483

    2016

  45. [53]

    Self-supervised interest point detection and description for fisheye and perspective images,

    M. Mera-Trujillo, S. Patel, Y . Gu, and G. Doretto, “Self-supervised interest point detection and description for fisheye and perspective images,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 6497–6506. 15

    2023

  46. [54]

    Sosnet: Second order similarity regularization for local descriptor learning,

    Y . Tian, X. Yu, B. Fan, F. Wu, H. Heijnen, and V . Balntas, “Sosnet: Second order similarity regularization for local descriptor learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 11 016–11 025

    2019

  47. [55]

    Magsac: marginalizing sample consensus,

    D. Barath, J. Matas, and J. Noskova, “Magsac: marginalizing sample consensus,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2019, pp. 10 197–10 205

    2019

  48. [56]

    Magsac++, a fast, reliable and accurate robust estimator,

    D. Barath, J. Noskova, M. Ivashechkin, and J. Matas, “Magsac++, a fast, reliable and accurate robust estimator,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 1304–1312

    2020

  49. [57]

    Clkn: Cascaded lucas- kanade networks for image alignment,

    C.-H. Chang, C.-N. Chou, and E. Y . Chang, “Clkn: Cascaded lucas- kanade networks for image alignment,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2017, pp. 2213–2221

    2017

  50. [58]

    Robust homography estimation via dual principal component pursuit,

    T. Ding, Y . Yang, Z. Zhu, D. P. Robinson, R. Vidal, L. Kneip, and M. C. Tsakiris, “Robust homography estimation via dual principal component pursuit,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 6080–6089

    2020

  51. [59]

    Gyroflow: gyroscope-guided unsupervised optical flow learning,

    H. Li, K. Luo, and S. Liu, “Gyroflow: gyroscope-guided unsupervised optical flow learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021, pp. 12 869–12 878

    2021

  52. [60]

    Asflow: Unsupervised optical flow learning with adaptive pyramid sampling,

    S. Liu, K. Luo, A. Luo, C. Wang, F. Meng, and B. Zeng, “Asflow: Unsupervised optical flow learning with adaptive pyramid sampling,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 32, no. 7, pp. 4282–4295, 2022

    2022

  53. [61]

    Realflow: Em-based realistic optical flow dataset generation from videos,

    Y . Han, K. Luo, A. Luo, J. Liu, H. Fan, G. Luo, and S. Liu, “Realflow: Em-based realistic optical flow dataset generation from videos,” in Proceedings of the European Conference on Computer Vision (ECCV) . Springer, 2022, pp. 288–305

    2022

  54. [62]

    Learning optical flow with kernel patch attention,

    A. Luo, F. Yang, X. Li, and S. Liu, “Learning optical flow with kernel patch attention,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 8906–8915

    2022

  55. [63]

    Pdc-net+: En- hanced probabilistic dense correspondence network,

    P. Truong, M. Danelljan, R. Timofte, and L. Van Gool, “Pdc-net+: En- hanced probabilistic dense correspondence network,” IEEE Transactions on Pattern Analysis and Machine Intelligence , 2023

    2023

  56. [64]

    Glu-net: Global-local uni- versal network for dense flow and correspondences,

    P. Truong, M. Danelljan, and R. Timofte, “Glu-net: Global-local uni- versal network for dense flow and correspondences,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 6258–6268

    2020

  57. [65]

    Deep homography estimation for dynamic scenes,

    H. Le, F. Liu, S. Zhang, and A. Agarwala, “Deep homography estimation for dynamic scenes,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 7652–7661

    2020

  58. [66]

    Semi-supervised deep large-baseline homography estimation with progressive equivalence con- straint,

    H. Jiang, H. Li, Y . Lu, S. Han, and S. Liu, “Semi-supervised deep large-baseline homography estimation with progressive equivalence con- straint,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 1, 2023, pp. 1024–1032

    2023

  59. [67]

    Compressed domain deep video super-resolution,

    P. Chen, W. Yang, M. Wang, L. Sun, K. Hu, and S. Wang, “Compressed domain deep video super-resolution,” IEEE Transactions on Image Processing, vol. 30, pp. 7156–7169, 2021

    2021

  60. [68]

    Real-time action recognition with deeply transferred motion vector cnns,

    B. Zhang, L. Wang, Z. Wang, Y . Qiao, and H. Wang, “Real-time action recognition with deeply transferred motion vector cnns,” IEEE Transactions on Image Processing, vol. 27, no. 5, pp. 2326–2339, 2018

    2018

  61. [69]

    Cpga: Coding priors-guided aggregation network for compressed video quality enhancement,

    Q. Zhu, J. Hao, Y . Ding, Y . Liu, Q. Mo, M. Sun, C. Zhou, and S. Zhu, “Cpga: Coding priors-guided aggregation network for compressed video quality enhancement,” arXiv preprint arXiv:2403.10362 , 2024

    2024 arXiv

  62. [70]

    Real time video object segmentation in compressed domain,

    Z. Tan, B. Liu, Q. Chu, H. Zhong, Y . Wu, W. Li, and N. Yu, “Real time video object segmentation in compressed domain,” IEEE Transactions on Circuits and Systems for Video Technology , vol. 31, no. 1, pp. 175– 188, 2020

    2020

  63. [71]

    Mvflow: Deep optical flow estimation of compressed videos with motion vector prior,

    S. Zhou, X. Jiang, W. Tan, R. He, and B. Yan, “Mvflow: Deep optical flow estimation of compressed videos with motion vector prior,” in Proceedings of the 31st ACM International Conference on Multimedia , 2023, pp. 1964–1974

    2023

  64. [72]

    Superglue: Learning feature matching with graph neural networks,

    P.-E. Sarlin, D. DeTone, T. Malisiewicz, and A. Rabinovich, “Superglue: Learning feature matching with graph neural networks,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, 2020, pp. 4938–4947

    2020

  65. [73]

    Loftr: Detector- free local feature matching with transformers,

    J. Sun, Z. Shen, Y . Wang, H. Bao, and X. Zhou, “Loftr: Detector- free local feature matching with transformers,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021, pp. 8922–8931

    2021

  66. [74]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” in Proc. ICLR, 2015. Yike Liu received the B.E. degrees from the Univer- sity of Electronic Science and Technology of China (UESTC), Chengdu, China, in 2020. Currently, he is a Ph.D. student at School of Inf...

    2015

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.