Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Restereo: Diffusion stereo video generation and restoration

T0 review · 3 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 A single video diffusion model can turn a low-resolution monocular video into a restored, consistent stereo pair.

desk verdict A genuinely new dual-branch recipe for joint stereo generation and restoration, with solid ablations but real-world generalization claims outrunning the in-domain synthetic metrics. read the letter →

arxiv 2506.06023 v1 pith:FDGJLORF submitted 2025-06-06 cs.CV

Xingchang Huang , Ashish Kumar Singh , Florian Dubost , Cristina Nader Vasconcelos , Sakar Khattar , Liang Shi , Christian Theobalt , Cengiz Oztireli
show 1 more author
Gurprit Singh
This is my paper
classification cs.CV
keywords stereovideogenerationrestorationdiffusionmodelssyntheticdatawarpedmaskconditioninglow-resolutionlatentviewconsistency
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 claims that one video diffusion model can take a low-resolution monocular video and produce both a restored left view and a consistent right view, simultaneously performing stereo generation and restoration. Existing stereo video methods assume high-quality input and only inpaint occluded regions, whereas this method fine-tunes the model on degraded synthetic stereo clips and conditions it on warped masks, so one model learns both restoration and stereo generation. The practical appeal is that training uses a relatively small synthetic dataset with ground-truth depth, avoiding expensive real stereo capture, and the method is reported to outperform StereoCrafter and StereoDiffusion on low-resolution inputs in LPIPS, CLIP consistency, and user ratings. The central claim is that this transfer from synthetic degraded data to real low-quality video works.

What carries the argument

The machinery is a fine-tuned latent video diffusion U-Net, built on the Stable Video Diffusion architecture and initialized from StereoCrafter, trained with two randomly sampled branches: a left-to-right branch whose conditioning input is a disparity-warped, degraded left-view latent concatenated with a resized binary warped mask, and a left-to-left branch whose conditioning is a degraded left-view latent with a zero mask. Both branches supervise with mean-square error against clean target latents, and only the temporal layers are trainable. Training data comes from Kubric-rendered ShapeNet objects with 458 environment maps, degraded with blurring, downsampling, Gaussian noise, and JPEG compression using fixed random seeds per video, following the Real-ESRGAN degradation recipe. During inference the same two branches denoise with the warped mask and zero mask, and histogram matching aligns brightness between the two output views.

What would settle it

Run the released model on real low-quality videos with content outside the 14 ShapeNet classes, such as human faces, text, and highly reflective materials, and compare the restored left view and generated right view against StereoCrafter with Real-ESRGAN preprocessing; visible inconsistency or no sharpness gain on these inputs would weaken the claim that the method applies generally to low-quality real-world videos, and the paper already reports such a failure for specular highlights.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that restoration and stereo generation do not need separate models: one diffusion U-Net, fine-tuned on degraded synthetic stereo videos and conditioned on a binary warped mask, can simultaneously sharpen the input left view and synthesize a consistent right view. The left-to-right branch inpaints the disoccluded regions marked by the warped mask, while the left-to-left branch with a zero mask teaches the network to restore the input itself. Because both branches share weights and are trained with the same degradation augmentation, the model learns to enhance and generate under a single consistent inference pass; the paper reports that the generated pairs are sharper and more view-consistent than StereoCrafter with Real-ESRGAN or FMA-Net preprocessing.

Load-bearing premise

The load-bearing premise is that the fine-tuning distribution, synthetic ShapeNet objects with simple materials plus Real-ESRGAN-style degradations, is representative enough of real-world low-quality monocular video that the model will restore and generate consistent stereo on real footage.

Editorial extensions

If this is right

  • A low-quality monocular video no longer needs a separate super-resolution pass before stereo conversion; the same model restores and generates both views.
  • Synthetic data with ground-truth depth is enough to fine-tune a stereo-video generator and restorer, avoiding large real-world stereo capture and stereo-matching pipelines.
  • The model produces sharper left and right views than StereoCrafter with Real-ESRGAN or FMA-Net preprocessing, with better LPIPS and CLIP view consistency.
  • Generation quality improves as the fine-tuning dataset grows from 10 to 958 videos, so scaling synthetic data diversity is a direct route to better results.
  • Because both branches share weights and run in one model, no additional restoration network or extra storage is needed at inference.

Reading between the lines

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

  • If the transfer assumption holds, the same mask-conditioned two-branch recipe could extend to other pairwise or multi-view tasks, such as generating consistent views at larger baselines or simultaneous denoising and novel-view synthesis from degraded input.
  • The reported failure on specular highlights suggests the limiting factor is material and lighting diversity in the synthetic data, not the architecture, so adding procedurally varied materials and lighting would be a direct, testable improvement.
  • Because the model currently processes 16-frame clips, applying it to long videos requires temporal chunking; testing consistency across chunk boundaries is an obvious next evaluation.
  • The fixed-seed degradation augmentation is likely load-bearing for temporal restoration consistency; ablating it with independent per-frame degradations would reveal how much of that consistency depends on this choice.
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

3 major / 5 minor

Summary. The paper proposes Restereo, a pipeline that fine-tunes a video diffusion model (initialized from StereoCrafter) on a small synthetic stereo dataset generated with Kubric/ShapeNet, adding degradation augmentation so that a single model can simultaneously restore a low-quality monocular input and generate a consistent stereo left/right video. The method trains both a left-to-right branch (conditioned on a warped occlusion mask) and a left-to-left restoration branch, and at inference runs both branches with histogram matching. Experiments compare against StereoCrafter and StereoDiffusion on synthetic and real videos, reporting LPIPS/CLIP consistency metrics, a small user study, and ablations on augmentation and training-data size.

Significance. If the central claims hold, this is a valuable contribution: a single model performing both stereo video generation and restoration, trained without large-scale real stereo video data. The idea of joint restoration and stereo generation with degradation augmentation is novel and practically useful, and the proposed use of synthetic data with ground-truth depth avoids the expensive preprocessing pipeline of prior work. The ablations on data size and augmentation are informative, and the commitment to release code and data is a strength. The main weaknesses are in evaluation: the primary quantitative generation-quality result is measured on a test set drawn from the same synthetic distribution as training, and the real-world evaluation does not verify stereo geometric correctness. These issues are addressable with additional experiments or with more carefully scoped claims.

major comments (3)
  1. [§4.3, Table 3] The quantitative claim of outperforming existing approaches on low-resolution inputs is based solely on LPIPS computed on 97 Kubric test videos generated with the same renderer, the same 14 ShapeNet classes, the same 458 environment maps, and the same degradation family used in training. This is an in-distribution test and does not support the abstract's assertion that the method 'can be ... applied to low-quality real-world videos.' The real-world evidence consists of qualitative figures (Figs. 4–5), consistency metrics (Table 1), and a small user study (Table 2), none of which compare generated right views to ground-truth stereo. To support the central claim, the paper should either add a quantitative evaluation on real stereo videos with ground truth (or a held-out dataset with a different distribution) or explicitly scope the quantitative claim to the synthetic distribution.
  2. [§4.3, Table 1] The view consistency metric LPIPS (and CLIP cosine similarity) measures perceptual similarity between left and right views, so a degenerate solution that outputs nearly identical views would score well even with zero disparity. The paper does not report any geometric validation (e.g., disparity consistency, epipolar error, or warping error against a known stereo pair), so the low LPIPS view score does not establish correct stereo geometry. Given the paper's own admission of failure on specular and highly reflective materials (Sec. 4.2, supplemental Fig. 6), a geometric check is needed to rule out plausible-but-incorrect stereo content.
  3. [§3.2 and §4.2] The fine-tuning data is limited to 14 ShapeNet object classes, simple materials, and degradation types from a Real-ESRGAN-style augmentation family; the paper acknowledges that this leads to failure on complex material appearance. This admitted limitation directly constrains the central claim of applicability to 'low-quality real-world videos,' which often contain diverse materials, textures, and scene types. The paper should either present evidence that the method handles a broader range of real-world content (e.g., a more diverse real-world test set with multiple scene categories) or relax the claim accordingly.
minor comments (5)
  1. [§4.1] The sentence 'All test videos are not unseen during training' is ambiguous; presumably it should read 'All test videos are unseen during training.' Please clarify, because the validity of the quantitative results depends on this.
  2. [§4.3, Table 2] The user study uses 15 participants, each scoring only 5 of 15 videos, and the standard deviations overlap between methods (e.g., 3D Stereo Effect 3.81±0.11 vs 4.07±0.10). A significance test or a larger participant pool would strengthen the claim of 'consistent improvements.'
  3. [§4.1 / §4.2] The paper compares only with StereoCrafter and StereoDiffusion, while the abstract claims to outperform 'existing approaches.' Since other training-based methods (SpatialMe, StereoConversion, ImmersePro) are discussed in the related work, adding at least one such comparison or qualifying the claim would be appropriate.
  4. [§3.3] The disparity scaling S and the morphological dilation kernel size for the warped mask are set without ablations; reporting sensitivity of these choices would help reproducibility and robustness assessment.
  5. [§4.1, Eq. (1)] The notation in Eq. (1) is a bit inconsistent, since x is used for both left and right views after saying superscripts are omitted; please clarify the definitions and the exact role of the 'Up' operation in the training versus testing settings.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper's claims are empirical, and the in-domain synthetic benchmark is a limitation on generalization evidence, not a construction-level reduction.

full rationale

The paper's load-bearing claims are empirical rather than derived. The model is fine-tuned on degraded synthetic Kubric/ShapeNet stereo videos and evaluated on (i) held-out Kubric test videos, (ii) real-world videos through qualitative comparisons and a user study, and (iii) LPIPS/CLIP view and temporal consistency metrics. None of these evaluations is equivalent by construction to the training signal: the held-out test videos use non-overlapping objects and environment maps, the metrics are computed on model outputs rather than being directly optimized, and the real-world qualitative and user-study evidence is independent of the Kubric fine-tuning distribution. The paper explicitly discloses the synthetic-data limitation, stating that 'Our method can fail to reproduce the details of specular highlight with highly reflected material ... as our data for fine-tuning do not contain objects with complex material appearance,' and it does not rely on any author-specific uniqueness theorem or self-citation chain to force its design. The observation that the main quantitative benchmark (Table 3) is in-distribution relative to fine-tuning is a legitimate limitation on how strongly the numbers support real-world transfer, but it is not a circular derivation: the comparison against StereoCrafter, Real-ESRGAN, and FMA-Net on the same degraded inputs remains a genuine empirical test. No circular step is present.

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

The central claim rests on the transfer of a synthetic-trained model to real video and on the quality of an external depth estimator; these are domain assumptions rather than free parameters of a derivation. The only explicit hand-set parameter in the presented pipeline is the disparity scale S, plus the unspecified degradation and mask postprocessing settings.

free parameters (3)
  • disparity scaling S at inference = 0.03
    Hand-set for all test videos; training searches S per video in [0.02, 0.2] to make the forward-warped left view overlap the right view. The fixed inference value converts depth from DepthCrafter to pixel disparity, so the perceived depth strength depends on this arbitrary choice. The paper claims robustness to S (supplemental Fig. 2) but does not quantify the operating range.
  • augmentation degradation severity = not specified (ranges from Real-ESRGAN implementation)
    The blur, noise, downsampling, and JPEG parameters are drawn from Wang et al. [45] ranges, but the exact distributions are not listed. These are hand-chosen and determine the degradation types the model can restore; a mismatch with real degradations would weaken the restoration claim.
  • morphological dilation kernel size for warped mask postprocessing = not specified
    The paper states that the warping mask is postprocessed using morphological dilation to eliminate flying pixels and holes, but the kernel size and number of iterations are not reported. This affects the inpainting region and can influence the right-view quality.
assumptions (4)
  • domain assumption DepthCrafter produces sufficiently accurate video depth for real-world monocular videos to yield useful forward-warped right-view masks.
    The pipeline relies on warped masks derived from estimated depth; if depth is inaccurate, the mask will have artifacts and the inpainted right view will be misaligned. The paper acknowledges depth quality affects the mask and applies morphological dilation, but provides no quantitative depth evaluation on real videos.
  • domain assumption The Kubric synthetic training distribution (ShapeNet 14 classes, 458 environment maps, no complex materials) is representative of real-world low-quality videos.
    Fine-tuning on 958 synthetic videos is supposed to transfer to real footage. The paper itself notes failure on specular and reflective materials (supplemental Fig. 6), indicating the assumption is only partially satisfied.
  • domain assumption The pretrained StereoCrafter weights provide a strong prior for stereo video generation that is preserved through fine-tuning.
    The authors initialize from StereoCrafter and only train temporal layers; if this prior did not exist or was not suitable, the fine-tuning results would be different. The method's success depends on this transfer, which is not tested from a different initialization.
  • domain assumption The Real-ESRGAN-style degradation used in training matches the degradations in real low-quality inputs.
    The method learns restoration from synthetic degradations; if real low-quality videos contain artifacts outside this family (for example, sensor noise correlations or unusual motion blur patterns), the restoration may fail. No real-world restoration metric is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Restereo: Diffusion stereo video generation and restoration." pith.science (2026). https://pith.science/paper/FDGJLORF

@misc{pith2026250606023,
  author       = {Pith},
  title        = {Pith review of: Restereo: Diffusion stereo video generation and restoration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FDGJLORF}},
  note         = {Machine review of arXiv:2506.06023}
}
read the original abstract

Stereo video generation has been gaining increasing attention with recent advancements in video diffusion models. However, most existing methods focus on generating 3D stereoscopic videos from monocular 2D videos. These approaches typically assume that the input monocular video is of high quality, making the task primarily about inpainting occluded regions in the warped video while preserving disoccluded areas. In this paper, we introduce a new pipeline that not only generates stereo videos but also enhances both left-view and right-view videos consistently with a single model. Our approach achieves this by fine-tuning the model on degraded data for restoration, as well as conditioning the model on warped masks for consistent stereo generation. As a result, our method can be fine-tuned on a relatively small synthetic stereo video datasets and applied to low-quality real-world videos, performing both stereo video generation and restoration. Experiments demonstrate that our method outperforms existing approaches both qualitatively and quantitatively in stereo video generation from low-resolution inputs.

Figures

Figures reproduced from arXiv: 2506.06023 by the authors.

Figure 1
Figure 1. Training and inference pipeline of our method. We fine-tune the Diffusion U-Net for both left-to-right and left-to-left generation [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Data augmentation with degradations is the key for restoration given low-resolution input. Our right-view output with augmentation [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The color histogram of Ours with histogram matching be [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Stereo generation comparisons between StereoDiffusion [ [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Stereo generation and restoration comparisons between StereoCrafter [ [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Elastic3D: Controllable Stereo Video Conversion with Guided Latent Decoding

    cs.CV 2025-12 conditional novelty 6.0 of 10

    Elastic3D converts monocular video to stereo by directly synthesizing the right-eye view with a one-step latent diffusion model conditioned on a user-set median disparity, using a guided decoder to preserve left-view details.

Reference graph

Works this paper leans on

56 extracted references · 32 canonical work pages · cited by 1 Pith paper

  1. [1]

    Immersity ai: The ai platform converting images and videos into 3d.https://www.immersity.ai/. 2

    work page

  2. [2]

    https: //www.owl3d.com/

    Owl3d: Ai-powered 2d to 3d conversion software. https: //www.owl3d.com/. 2

    work page

  3. [3]

    Stable video diffusion: Scaling latent video diffusion models to large datasets.arXiv preprint arXiv:2311.15127, 2023

    Andreas Blattmann, Tim Dockhorn, Sumith Kulal, Daniel Mendelevitch, Maciej Kilian, Dominik Lorenz, Yam Levi, Zion English, Vikram V oleti, Adam Letts, et al. Stable video diffusion: Scaling latent video diffusion models to large datasets.arXiv preprint arXiv:2311.15127, 2023. 2, 4, 5

    arXiv 2023

  4. [4]

    Blender Foundation. Blender. https://www.blender. org, 2023. Version 3.4. 3 9

    work page 2023

  5. [5]

    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. InComputer Vision–ECCV 2012: 12th European Conference on Computer Vision, Florence, Italy, October 7- 13, 2012, Proceedings, Part VI 12, pages 611–625. Springer,

    work page 2012

  6. [6]

    Basicvsr: The search for essential compo- nents in video super-resolution and beyond

    Kelvin CK Chan, Xintao Wang, Ke Yu, Chao Dong, and Chen Change Loy. Basicvsr: The search for essential compo- nents in video super-resolution and beyond. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4947–4956, 2021. 2

    work page 2021

  7. [7]

    Basicvsr++: Improving video super- resolution with enhanced propagation and alignment

    Kelvin CK Chan, Shangchen Zhou, Xiangyu Xu, and Chen Change Loy. Basicvsr++: Improving video super- resolution with enhanced propagation and alignment. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 5972–5981, 2022. 2

    work page 2022

  8. [8]

    Shapenet: An information- rich 3d model repository.arXiv preprint arXiv:1512.03012,

    Angel X Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese, Manolis Savva, Shuran Song, Hao Su, et al. Shapenet: An information- rich 3d model repository.arXiv preprint arXiv:1512.03012,

    arXiv

Show all 56 references
  1. [9]

    Svg: 3d stereoscopic video generation via denoising frame matrix

    Peng Dai, Feitong Tan, Qiangeng Xu, David Futschik, Ruofei Du, Sean Fanello, Xiaojuan Qi, and Yinda Zhang. Svg: 3d stereoscopic video generation via denoising frame matrix. arXiv preprint arXiv:2407.00367, 2024. 1, 2, 9

    2024 arXiv

  2. [10]

    Objaverse: A universe of annotated 3d objects

    Matt Deitke, Dustin Schwenk, Jordi Salvador, Luca Weihs, Oscar Michel, Eli VanderBilt, Ludwig Schmidt, Kiana Ehsani, Aniruddha Kembhavi, and Ali Farhadi. Objaverse: A universe of annotated 3d objects. InProceedings of the IEEE/CVF conference on computer vision and pattern reco...

    2023

  3. [11]

    Cat3d: Create anything in 3d with multi-view diffusion models.arXiv preprint arXiv:2405.10314, 2024

    Ruiqi Gao, Aleksander Holynski, Philipp Henzler, Arthur Brussee, Ricardo Martin-Brualla, Pratul Srinivasan, Jonathan T Barron, and Ben Poole. Cat3d: Create anything in 3d with multi-view diffusion models.arXiv preprint arXiv:2405.10314, 2024. 2

    2024 arXiv

  4. [12]

    Kubric: A scalable dataset generator

    Klaus Greff, Francois Belletti, Lucas Beyer, Carl Doersch, Yilun Du, Daniel Duckworth, David J Fleet, Dan Gnanapra- gasam, Florian Golemo, Charles Herrmann, et al. Kubric: A scalable dataset generator. InProceedings of the IEEE/CVF conference on computer vision and pattern rec...

    2022

  5. [13]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bern- hard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilibrium. Advances in neural information processing systems, 30, 2017. 9

    2017

  6. [14]

    Deep sky modeling for single image out- door lighting estimation

    Yannick Hold-Geoffroy, Akshaya Athawale, and Jean- Franc ¸ois Lalonde. Deep sky modeling for single image out- door lighting estimation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 6927–6935, 2019. 3

    2019

  7. [15]

    Training-free camera control for video generation.arXiv preprint arXiv:2406.10126, 2024

    Chen Hou, Guoqiang Wei, Yan Zeng, and Zhibo Chen. Training-free camera control for video generation.arXiv preprint arXiv:2406.10126, 2024. 1

    2024 arXiv

  8. [16]

    Depthcrafter: Generating consistent long depth sequences for open-world videos.arXiv preprint arXiv:2409.02095, 2024

    Wenbo Hu, Xiangjun Gao, Xiaoyu Li, Sijie Zhao, Xiaodong Cun, Yong Zhang, Long Quan, and Ying Shan. Depthcrafter: Generating consistent long depth sequences for open-world videos.arXiv preprint arXiv:2409.02095, 2024. 4, 5, 6

    2024 arXiv

  9. [17]

    T-svg: Text-driven stereoscopic video generation

    Qiao Jin, Xiaodong Chen, Wu Liu, Tao Mei, and Yongdong Zhang. T-svg: Text-driven stereoscopic video generation. arXiv preprint arXiv:2412.09323, 2024. 1

    2024 arXiv

  10. [18]

    Clevr: A diagnostic dataset for compositional language and elementary visual reasoning

    Justin Johnson, Bharath Hariharan, Laurens Van Der Maaten, Li Fei-Fei, C Lawrence Zitnick, and Ross Girshick. Clevr: A diagnostic dataset for compositional language and elementary visual reasoning. InProceedings of the IEEE conference on computer vision and pattern recognition...

  11. [19]

    Elucidating the design space of diffusion-based generative models.Advances in neural information processing systems, 35:26565–26577, 2022

    Tero Karras, Miika Aittala, Timo Aila, and Samuli Laine. Elucidating the design space of diffusion-based generative models.Advances in neural information processing systems, 35:26565–26577, 2022. 6

    2022

  12. [20]

    3d gaussian splatting for real-time radiance field rendering.ACM Trans

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk¨uhler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering.ACM Trans. Graph., 42(4):139–1, 2023. 2

    2023

  13. [21]

    Trackdif- fusion: Multi-object tracking data generation via diffusion models.arXiv preprint arXiv:2312.00651, 2023

    Pengxiang Li, Zhili Liu, Kai Chen, Lanqing Hong, Yunzhi Zhuge, Dit-Yan Yeung, Huchuan Lu, and Xu Jia. Trackdif- fusion: Multi-object tracking data generation via diffusion models.arXiv preprint arXiv:2312.00651, 2023. 6

    2023 arXiv

  14. [22]

    Towards an end-to-end framework for flow- guided video inpainting

    Zhen Li, Cheng-Ze Lu, Jianhua Qin, Chun-Le Guo, and Ming- Ming Cheng. Towards an end-to-end framework for flow- guided video inpainting. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 17562–17571, 2022. 2

    2022

  15. [23]

    Dynibar: Neural dynamic image-based rendering

    Zhengqi Li, Qianqian Wang, Forrester Cole, Richard Tucker, and Noah Snavely. Dynibar: Neural dynamic image-based rendering. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4273–4284,

  16. [24]

    Recurrent video restoration trans- former with guided deformable attention.Advances in Neural Information Processing Systems, 35:378–393, 2022

    Jingyun Liang, Yuchen Fan, Xiaoyu Xiang, Rakesh Ranjan, Eddy Ilg, Simon Green, Jiezhang Cao, Kai Zhang, Radu Timofte, and Luc V Gool. Recurrent video restoration trans- former with guided deformable attention.Advances in Neural Information Processing Systems, 35:378–393, 2022. 2

    2022

  17. [25]

    Vrt: A video restoration transformer.IEEE Transactions on Image Processing, 2024

    Jingyun Liang, Jiezhang Cao, Yuchen Fan, Kai Zhang, Rakesh Ranjan, Yawei Li, Radu Timofte, and Luc Van Gool. Vrt: A video restoration transformer.IEEE Transactions on Image Processing, 2024. 2

    2024

  18. [26]

    Diffbir: Toward blind image restoration with generative diffusion prior

    Xinqi Lin, Jingwen He, Ziyan Chen, Zhaoyang Lyu, Bo Dai, Fanghua Yu, Yu Qiao, Wanli Ouyang, and Chao Dong. Diffbir: Toward blind image restoration with generative diffusion prior. InEuropean Conference on Computer Vision, pages 430–448. Springer, 2024. 2

    2024

  19. [27]

    Robust dynamic radiance fields

    Yu-Lun Liu, Chen Gao, Andreas Meuleman, Hung-Yu Tseng, Ayush Saraf, Changil Kim, Yung-Yu Chuang, Johannes Kopf, and Jia-Bin Huang. Robust dynamic radiance fields. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13–23, 2023. 2

    2023

  20. [28]

    Decoupled weight decay regularization.arXiv preprint arXiv:1711.05101, 2017

    I Loshchilov. Decoupled weight decay regularization.arXiv preprint arXiv:1711.05101, 2017. 6

    2017 arXiv

  21. [29]

    Repaint: Inpainting us- ing denoising diffusion probabilistic models

    Andreas Lugmayr, Martin Danelljan, Andres Romero, Fisher Yu, Radu Timofte, and Luc Van Gool. Repaint: Inpainting us- ing denoising diffusion probabilistic models. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11461–11471, 2022. 2 10

    2022

  22. [30]

    Stereo conversion with disparity-aware warping, compositing and inpainting

    Lukas Mehl, Andr´es Bruhn, Markus Gross, and Christopher Schroers. Stereo conversion with disparity-aware warping, compositing and inpainting. InProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 4260–4269, 2024. 1, 2

    2024

  23. [31]

    Object scene flow for au- tonomous vehicles

    Moritz Menze and Andreas Geiger. Object scene flow for au- tonomous vehicles. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 3061–3070,

  24. [32]

    Nerf: Representing scenes as neural radiance fields for view syn- thesis.Communications of the ACM, 65(1):99–106, 2021

    Ben Mildenhall, Pratul P Srinivasan, Matthew Tancik, Jonathan T Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view syn- thesis.Communications of the ACM, 65(1):99–106, 2021. 2

    2021

  25. [33]

    Pytorch: An imperative style, high-performance deep learning library.Ad- vances in neural information processing systems, 32, 2019

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. Pytorch: An imperative style, high-performance deep learning library.Ad- vances in neural information processing systems, 32...

    2019

  26. [34]

    Pixabay license summary

    Pixabay. Pixabay license summary. https://pixabay. com/service/license-summary/, 2025. Accessed: 2025-03-03. 4, 5, 6

    2025

  27. [35]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. InInternational conference on machine learning, p...

    2021

  28. [36]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022. 1, 6

    2022

  29. [37]

    U- net: Convolutional networks for biomedical image segmen- tation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U- net: Convolutional networks for biomedical image segmen- tation. InMedical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, par...

    2015

  30. [38]

    Immersepro: End- to-end stereo video synthesis via implicit disparity learning

    Jian Shi, Zhenyu Li, and Peter Wonka. Immersepro: End- to-end stereo video synthesis via implicit disparity learning. arXiv preprint arXiv:2410.00262, 2024. 1, 2

    2024 arXiv

  31. [39]

    Stereocrafter-zero: Zero-shot stereo video generation with noisy restart.arXiv preprint arXiv:2411.14295, 2024

    Jian Shi, Qian Wang, Zhenyu Li, and Peter Wonka. Stereocrafter-zero: Zero-shot stereo video generation with noisy restart.arXiv preprint arXiv:2411.14295, 2024. 1

    2024 arXiv

  32. [40]

    Fvd: A new metric for video generation

    Thomas Unterthiner, Sjoerd Van Steenkiste, Karol Kurach, Rapha¨el Marinier, Marcin Michalski, and Sylvain Gelly. Fvd: A new metric for video generation. 2019. 9

    2019

  33. [41]

    Sch¨onberger, Juan Nunez- Iglesias, Franc ¸ois Boulogne, Joshua D

    St´efan van der Walt, Johannes L. Sch¨onberger, Juan Nunez- Iglesias, Franc ¸ois Boulogne, Joshua D. Warner, Neil Yager, Emmanuelle Gouillart, Tony Yu, and the scikit-image con- tributors. scikit-image: image processing in Python.PeerJ, 2: e453, 2014. 5

    2014

  34. [42]

    Generative camera dolly: Extreme monoc- ular dynamic novel view synthesis

    Basile Van Hoorick, Rundi Wu, Ege Ozguroglu, Kyle Sargent, Ruoshi Liu, Pavel Tokmakov, Achal Dave, Changxi Zheng, and Carl V ondrick. Generative camera dolly: Extreme monoc- ular dynamic novel view synthesis. InEuropean Conference on Computer Vision, pages 313–331. Springer, 2024. 2

    2024

  35. [43]

    Modelscope text-to-video technical report.arXiv preprint arXiv:2308.06571, 2023

    Jiuniu Wang, Hangjie Yuan, Dayou Chen, Yingya Zhang, Xiang Wang, and Shiwei Zhang. Modelscope text-to-video technical report.arXiv preprint arXiv:2308.06571, 2023. 1

    2023 arXiv

  36. [44]

    Stereodiffusion: Training-free stereo image generation using latent diffusion models

    Lezhong Wang, Jeppe Revall Frisvad, Mark Bo Jensen, and Siavash Arjomand Bigdeli. Stereodiffusion: Training-free stereo image generation using latent diffusion models. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7416–7425, 2024....

    2024

  37. [45]

    Real-esrgan: Training real-world blind super-resolution with pure synthetic data

    Xintao Wang, Liangbin Xie, Chao Dong, and Ying Shan. Real-esrgan: Training real-world blind super-resolution with pure synthetic data. InProceedings of the IEEE/CVF inter- national conference on computer vision, pages 1905–1914,

    1905

  38. [46]

    Cat4d: Create anything in 4d with multi-view video diffusion models

    Rundi Wu, Ruiqi Gao, Ben Poole, Alex Trevithick, Changxi Zheng, Jonathan T Barron, and Aleksander Holynski. Cat4d: Create anything in 4d with multi-view video diffusion models. arXiv preprint arXiv:2411.18613, 2024. 2

    2024 arXiv

  39. [47]

    Sv4d: Dynamic 3d content generation with multi-frame and multi-view consistency.arXiv preprint arXiv:2407.17470, 2024

    Yiming Xie, Chun-Han Yao, Vikram V oleti, Huaizu Jiang, and Varun Jampani. Sv4d: Dynamic 3d content generation with multi-frame and multi-view consistency.arXiv preprint arXiv:2407.17470, 2024. 2

    2024 arXiv

  40. [48]

    Drivingstereo: A large-scale dataset for stereo matching in autonomous driving scenarios

    Guorun Yang, Xiao Song, Chaoqin Huang, Zhidong Deng, Jianping Shi, and Bolei Zhou. Drivingstereo: A large-scale dataset for stereo matching in autonomous driving scenarios. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 899–908, 2019. 4

    2019

  41. [49]

    Diffir2vr-zero: Zero-shot video restoration with diffusion-based image restoration models.arXiv preprint arXiv:2407.01519, 2024

    Chang-Han Yeh, Chin-Yang Lin, Zhixiang Wang, Chi- Wei Hsiao, Ting-Hsuan Chen, Hau-Shiang Shiu, and Yu- Lun Liu. Diffir2vr-zero: Zero-shot video restoration with diffusion-based image restoration models.arXiv preprint arXiv:2407.01519, 2024. 2

    2024

  42. [50]

    Fma-net: Flow-guided dynamic filtering and iterative feature refine- ment with multi-attention for joint video super-resolution and deblurring

    Geunhyuk Youk, Jihyong Oh, and Munchurl Kim. Fma-net: Flow-guided dynamic filtering and iterative feature refine- ment with multi-attention for joint video super-resolution and deblurring. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pag...

    2024

  43. [51]

    Spatialme: Stereo video conversion us- ing depth-warping and blend-inpainting.arXiv preprint arXiv:2412.11512, 2024

    Jiale Zhang, Qianxi Jia, Yang Liu, Wei Zhang, Wei Wei, and Xin Tian. Spatialme: Stereo video conversion us- ing depth-warping and blend-inpainting.arXiv preprint arXiv:2412.11512, 2024. 1, 2

    2024 arXiv

  44. [52]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 586–595, 2018. 9

    2018

  45. [53]

    Stereocrafter: Diffusion-based generation of long and high-fidelity stereoscopic 3d from monocular videos.arXiv preprint arXiv:2409.07447, 2024

    Sijie Zhao, Wenbo Hu, Xiaodong Cun, Yong Zhang, Xi- aoyu Li, Zhe Kong, Xiangjun Gao, Muyao Niu, and Ying Shan. Stereocrafter: Diffusion-based generation of long and high-fidelity stereoscopic 3d from monocular videos.arXiv preprint arXiv:2409.07447, 2024. 1, 2, 3, 4, 5, 6, 7, 8

    2024 arXiv

  46. [54]

    clip-score: CLIP Score for Py- Torch

    SUN Zhengwentai. clip-score: CLIP Score for Py- Torch. https : / / github . com / taited / clip - score, 2023. Version 0.1.1. 9 11

    2023

  47. [55]

    ProPainter: Improving propagation and transformer for video inpainting

    Shangchen Zhou, Chongyi Li, Kelvin C.K Chan, and Chen Change Loy. ProPainter: Improving propagation and transformer for video inpainting. InProceedings of IEEE International Conference on Computer Vision (ICCV), 2023. 2

    2023

  48. [56]

    Upscale-a-video: Temporal-consistent diffusion model for real-world video super-resolution

    Shangchen Zhou, Peiqing Yang, Jianyi Wang, Yihang Luo, and Chen Change Loy. Upscale-a-video: Temporal-consistent diffusion model for real-world video super-resolution. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2535–2545, 2024. 2 12

    2024

Pith tools

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