Pith. sign in

REVIEW 4 major objections 6 minor 36 references

Audio-Assisted Face Video Restoration with Temporal and Identity Complementary Learning

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read GAVN claims to outdo the state of the art in face video restoration on compression artifact removal, deblurring, and super-resolution by combining inter-frame temporal features with intra-frame identity features extracted from audio and lan

desk verdict A plausible audio-assisted face video restoration method that handles three degradations, but the central SOTA claim is compromised by an obvious Table 1 typo and missing code/reproducibility. read the letter →

arxiv 2508.04161 v1 pith:GL6ATZNM submitted 2025-08-06 cs.CV cs.MMcs.SDeess.AS

Yuqin Cao , Yixuan Gao , Wei Sun , Xiaohong Liu , Yulun Zhang , Xiongkuo Min This is my paper · ORCID
classification cs.CVcs.MMcs.SDeess.AS
keywords audio-assistedvideorestorationfacesuper-resolutioncompressionartifactremovaldeblurringfaciallandmarkdetectionlip-syncdeformableconvolution
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that audio, which streaming video already carries, can serve as a general-purpose restoration prior for faces, not just for compression artifacts as in earlier audio-aided work, but for deblurring and super-resolution too. GAVN splits the problem: a temporal branch aligns and fuses neighboring frames in a cheap low-resolution space to restore coarse motion structure, while an identity branch fuses the current high-resolution frame with audio features and face landmarks to recover fine facial detail, especially in the mouth and eyes. Trained on a multi-speaker dataset (VoxCeleb2) and a single-speaker dataset (Obama), GAVN reports the best PSNR, SSIM, MS-SSIM, LPIPS, and lip-sync scores against five video-restoration baselines across all three degradation types. The paper's ablations indicate both added branches matter: removing identity features drops quality to roughly BasicVSR++ level, and removing the audio features hurts both image quality and lip-sync consistency. If the result holds, audio becomes a near-free side channel for improving degraded talking-face video.

What carries the argument

The load-bearing mechanism is the two-branch feature architecture with a deliberate division of labor. The inter-frame temporal module applies deformable-convolution alignment, adjacent and skip-frame, forward and backward, followed by attention-weighted fusion in a low-resolution pyramid, providing coarse motion features at low computational cost. The intra-frame identity module works at full resolution on the single current frame, fusing frame features with audio features (from a Bi-LSTM) and landmark features (from a PFLD detector retrained on distorted frames with audio input) via attention-map weighting. The reconstruction module upsamples and merges the two feature streams. What carrie

What would settle it

Run GAVN on degraded frames whose audio track has been time-shifted or replaced with a different speaker's voice. If PSNR, SSIM, and SyncNet scores stay high, the audio branch is not the causal source of the improvement. Separately, measure the retrained versus original PFLD landmark error on the distorted validation frames; if the retrained detector is no more accurate, the identity module's benefit is not explained by better landmarks.

Watch

Extended reading notes

Core claim

The paper's central claim is that face video restoration improves when the network sees two complementary representations: a temporal one assembled from several consecutive frames aligned by deformable convolutions in downsampled space, and an identity one assembled from a single high-resolution frame, its audio segment, and its face landmarks. The audio signal is the enabling prior, because speech is physically shaped by the lips and mouth muscles, so the sound track carries information about precisely the regions that degrade most. GAVN reports the best results on both datasets across compression artifact removal, deblurring, and super-resolution, with qualitatively sharper eye contours an

Load-bearing premise

The method depends on the audio track and the retrained landmark detector actually supplying recoverable information about facial detail in degraded frames; the paper does not independently verify that the audio-conditioned retrained PFLD detects landmarks more accurately than the original on distorted frames, so part of the identity branch's measured gain could come from the landmark inputs themselves rather than from a validated identity prior.

Editorial extensions

If this is right

  • Audio-assisted restoration extends beyond compression artifact removal: the same network architecture handles deblurring and super-resolution, so streaming pipelines can use one audio-conditioned model for mixed degradations.
  • Restoring with identity features preserves speaker-specific appearance, which the paper argues matters because the face is a structured personal identifier, in particular when restored video feeds face recognition or verification.
  • Because the identity branch uses the current frame's audio and landmarks at full resolution, the gains concentrate where audio-visual correlation is strongest, the mouth region, with improved SyncNet lip-sync confidence on both synthetic and real-world degraded videos.
  • Retraining the landmark detector on distorted frames with audio input is claimed to make landmark priors usable on low-quality streaming video, a step beyond detectors trained only on clean faces.
  • On real-world YouTube videos with no ground truth, GAVN still reports the best NIQE naturalness and best SyncNet scores among compared methods, suggesting the gains are not an artifact of synthetic degradations.

Reading between the lines

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

  • A direct causal check the paper does not run: swap or time-shift the audio segment at inference. If restoration quality and lip-sync metrics barely change, the audio branch would be contributing associatively rather than through the claimed physical lip-sound coupling.
  • The same audio-conditioned, identity-preserving pipeline could extend to neighbouring problems where mouth-region fidelity matters, such as talking-head generation, audio-visual speech enhancement, or restorations feeding automatic speechreading, since the network already learns a lip-motion-to-speech mapping.
  • The paper motivates but does not measure computational cost; the low-resolution temporal branch is positioned for streaming use, so a runtime-versus-quality comparison against recurrent baselines such as BasicVSR++ and VRT would test that positioning.
  • The trick of retraining a landmark detector on degraded frames under audio guidance could generalize to other facial priors such as face parsing maps or 3D morphable-model coefficients, and to audio-conditioned detection beyond landmarks.
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 / 6 minor

Summary. The paper proposes GAVN, a general audio-assisted face video restoration network for compression artifact removal, deblurring, and super-resolution. GAVN uses an inter-frame temporal module operating in low-resolution space to capture motion and an intra-frame identity module operating in high-resolution space to capture facial details with the aid of audio features and face landmarks from a retrained PFLD detector. The two types of features are fused in a reconstruction module that produces high-quality frames. The method is evaluated on VoxCeleb2 and Obama datasets against DBPN, EDVR, BasicVSR++, DAVD-Net, and VRT, with additional experiments at multiple distortion levels, ablation studies, and a small real-world video set. The central claim is that GAVN outperforms existing state-of-the-art methods across all three restoration tasks.

Significance. If the reported results are reliable, the paper makes a useful contribution by extending audio-assisted face restoration beyond compression artifact removal to deblurring and super-resolution. The design of combining low-resolution temporal features with high-resolution identity features under audio guidance is reasonable, and the comparison retrains all baselines on the same data, which strengthens the fairness of the evaluation. The ablation studies separating identity and audio contributions are helpful. However, the empirical claims rest on single-run quantitative tables with at least one clear data-entry error, no variance or significance information, and no released code or models. The retrained landmark detector is also insufficiently described and verified. These issues need to be addressed before the SOTA claim can be fully accepted.

major comments (4)
  1. [Comparison with SOTA Methods, Table 1] In the Deblur block for VoxCeleb2, the VRT row reports Syncd = 0.7442, while every other Syncd value in that column is around 7.4. This is almost certainly a decimal-point typo (likely 7.442). Since Table 1 is the primary evidence for the paper's SOTA claim, this uncorrected error undermines confidence in the numerical reporting. Please correct the entry and carefully re-check all values in Tables 1, 2, and 3.
  2. [Training Details / Tables 1–3] All quantitative results appear to come from a single training run with no error bars, multiple seeds, or significance tests. Many of the reported margins are small (e.g., +0.11 to +0.44 dB PSNR in Table 2, and LPIPS differences near 0.004–0.006 in Table 1). Without variance information, it is not possible to determine whether these margins exceed seed noise. Please report mean and standard deviation over at least three runs, or provide significance tests, for the main comparisons and ablations.
  3. [Intra-Frame Identity Module] The paper states that PFLD is pretrained on the training set using distorted face frames and corresponding audio segments as input, with landmarks from original frames as ground truth. However, no architecture details are given for how audio is injected into PFLD, and no quantitative evaluation is provided to show that this retrained detector is more accurate on degraded frames than the original PFLD. Since the landmark features are a core component of the proposed identity module, the contribution cannot be clearly attributed. Please include landmark accuracy metrics (e.g., NME) on distorted frames, and an ablation comparing the retrained detector with the original detector.
  4. [Experiments on Real-World Degraded Face Videos / Table 4] The real-world evaluation uses only 10 videos and no-reference metrics. The reported improvements are very small (e.g., NIQE 6.1549 vs. 6.1620 for the closest competitor, and Syncd 7.1253 vs. 7.1657). With 10 videos, these differences may not be statistically meaningful. Please either enlarge the real-world set, provide per-video results with statistical testing, or temper the claim that GAVN 'outperforms' other methods on real-world data.
minor comments (6)
  1. [Throughout] The method name is inconsistently spaced as 'GA VN' in the abstract and body text but 'GAVN' in tables and figures. Please unify the notation.
  2. [Conclusion] The conclusion contains 'The integration of audio and identify features', where 'identify' should be 'identity'.
  3. [Figure 3(b) caption] The caption spells 'Indentity Module' instead of 'Identity Module'.
  4. [Figure 2] The section references in the caption appear as 'Sec.' without numbers; please fill in the proper section references.
  5. [Dataset names] The dataset names are inconsistently rendered as 'V oxCeleb2' and 'Obama dataset'; use the standard 'VoxCeleb2' and specify the exact Obama dataset version/release used.
  6. [Evaluation Criteria] The paper does not state whether the SyncNet evaluation uses the same audio features or window sizes across methods. Since GAVN is audio-assisted, please clarify the evaluation protocol to avoid any perceived bias.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; GAVN's claims are empirical and its components are trained against external ground truth, not against the evaluation metrics.

full rationale

GAVN is an empirical learning method, not a closed-form derivation, and no step in the paper reduces a predicted quantity to a fitted input by construction. The intra-frame identity module uses PFLD pretrained on distorted frames plus audio, with landmark detections from original frames as ground truth; this is auxiliary supervised pretraining, and the final restoration loss is a Charbonnier penalty against ground-truth high-quality frames, so the landmarks are not fitted to the reported PSNR/SSIM/SyncNet outcomes. DAVD-Net is cited for attention-map inspiration, but it is not by the current authors and is not load-bearing; no uniqueness theorem or ansatz is imported via self-citation. SyncNet is used only as an evaluation metric, not as a training loss, so using audio as an input does not make the metric circular—though it may raise a separate evaluation-bias concern. The apparent typo in Table 1 (VRT Syncd = 0.7442) and the lack of released code or multi-run statistics are correctness/reproducibility issues, not circularity. Overall, the paper's central claims are supported by held-out comparisons against external baselines and are not equivalent to its inputs by definition.

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

This paper is an empirical deep learning contribution; it introduces no new fitted constants or invented entities. The central claim rests on standard assumptions about data representation and metric validity, plus the specific belief that audio provides useful supervision for lip regions.

assumptions (4)
  • domain assumption Audio and lip movements are strongly correlated in face videos, so audio features can guide restoration of mouth regions.
    Introduced in the Introduction paragraph on motivation; the entire design of the intra-frame identity module depends on this correlation holding for the test distribution.
  • domain assumption The retrained PFLD landmark detector, trained on distorted frames with audio assistance, produces landmarks accurate enough to aid restoration.
    Methodology section on Intra-Frame Identity Module states this pretraining and relies on it for identity features.
  • domain assumption Existing degradation models (CRF 45 compression, Gaussian blur, bicubic downsampling) adequately represent real-world distortions for evaluation.
    Dataset section describes synthetic degradations; generalization to real-world is only tested on 10 YouTube videos with no ground truth.
  • domain assumption SyncNet confidence and distance are valid metrics for audio-visual synchronization quality.
    Evaluation criteria section uses SyncNet; the method directly uses audio, which may bias this metric.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Audio-Assisted Face Video Restoration with Temporal and Identity Complementary Learning." pith.science (2026). https://pith.science/paper/GL6ATZNM

@misc{pith2026250804161,
  author       = {Pith},
  title        = {Pith review of: Audio-Assisted Face Video Restoration with Temporal and Identity Complementary Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GL6ATZNM}},
  note         = {Machine review of arXiv:2508.04161}
}
read the original abstract

Face videos accompanied by audio have become integral to our daily lives, while they often suffer from complex degradations. Most face video restoration methods neglect the intrinsic correlations between the visual and audio features, especially in mouth regions. A few audio-aided face video restoration methods have been proposed, but they only focus on compression artifact removal. In this paper, we propose a General Audio-assisted face Video restoration Network (GAVN) to address various types of streaming video distortions via identity and temporal complementary learning. Specifically, GAVN first captures inter-frame temporal features in the low-resolution space to restore frames coarsely and save computational cost. Then, GAVN extracts intra-frame identity features in the high-resolution space with the assistance of audio signals and face landmarks to restore more facial details. Finally, the reconstruction module integrates temporal features and identity features to generate high-quality face videos. Experimental results demonstrate that GAVN outperforms the existing state-of-the-art methods on face video compression artifact removal, deblurring, and super-resolution. Codes will be released upon publication.

Figures

Figures reproduced from arXiv: 2508.04161 by the authors.

Figure 1
Figure 1. Restoration results of our proposed GAVN with [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The framework of the proposed GAVN, which consists of three modules: (a) Inter-Frame Temporal Module (Sec. ): [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Details of the reconstruction module. It inte [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (3 more)
Figure 5
Figure 5. Figure 5: Qualitative results on VoxCeleb2 dataset. Distortion types from top to bottom: compression, blur, and low resolution. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Qualitative results on the Obama dataset. Distortion types from top to bottom: compression, blur, and low resolution. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Qualitative results on real-world videos. [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 30 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Bulat, A.; and Tzimiropoulos, G. 2018. Super-fan: Integrated facial landmark localization and super-resolution of real-world low resolution faces in arbitrary poses with gans. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 109--117

    work page 2018

  4. [4]

    C.; Wang, X.; Yu, K.; Dong, C.; and Loy, C

    Chan, K. C.; Wang, X.; Yu, K.; Dong, C.; and Loy, C. C. 2021. Basicvsr: The search for essential components in video super-resolution and beyond. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 4947--4956

    work page 2021

  5. [5]

    C.; Zhou, S.; Xu, X.; and Loy, C

    Chan, K. C.; Zhou, S.; Xu, X.; and Loy, C. C. 2022. Basicvsr++: Improving video super-resolution with enhanced propagation and alignment. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 5972--5981

    work page 2022

  6. [6]

    Chen, C.; Seff, A.; Kornhauser, A.; and Xiao, J. 2015. Deepdriving: Learning affordance for direct perception in autonomous driving. In Proceedings of the IEEE International Conference on Computer Vision, 2722--2730

    work page 2015

  7. [7]

    Chen, Y.; Tai, Y.; Liu, X.; Shen, C.; and Yang, J. 2018. Fsrnet: End-to-end learning face super-resolution with facial priors. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2492--2501

    work page 2018

  8. [8]

    S.; Nagrani, A.; and Zisserman, A

    Chung, J. S.; Nagrani, A.; and Zisserman, A. 2018. Voxceleb2: Deep speaker recognition. arXiv preprint arXiv:1806.05622

    arXiv 2018

Show all 36 references
  1. [9]

    S.; and Zisserman, A

    Chung, J. S.; and Zisserman, A. 2017. Out of time: automated lip sync in the wild. In ACCV, 251--263

    2017

  2. [10]

    Guo, S.; Yang, X.; Ma, J.; Ren, G.; and Zhang, L. 2022. A differentiable two-stage alignment scheme for burst image reconstruction with large shift. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 17472--17481

    2022

  3. [11]

    Guo, X.; Li, S.; Yu, J.; Zhang, J.; Ma, J.; Ma, L.; Liu, W.; and Ling, H. 2019. PFLD : A practical facial landmark detector. arXiv preprint arXiv:1902.10859

    2019 arXiv

  4. [12]

    Guo, Y.; Zhang, X.; and Wu, X. 2020. Deep multi-modality soft-decoding of very low bit-rate face videos. In Proceedings of the ACM International Conference on Multimedia, 3947--3955

    2020

  5. [13]

    Haris, M.; Shakhnarovich, G.; and Ukita, N. 2018. Deep back-projection networks for super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 1664--1673

    2018

  6. [14]

    Haris, M.; Shakhnarovich, G.; and Ukita, N. 2019. Recurrent back-projection network for video super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 3897--3906

    2019

  7. [15]

    Isobe, T.; Li, S.; Jia, X.; Yuan, S.; Slabaugh, G.; Xu, C.; Li, Y.-L.; Wang, S.; and Tian, Q. 2020. Video super-resolution with temporal group attention. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 8008--8017

    2020

  8. [16]

    Ketkar, N.; Moolayil, J.; Ketkar, N.; and Moolayil, J. 2021. Introduction to pytorch. Deep Learning with Python: Learn Best Practices of Deep Learning Models with PyTorch, 27--91

    2021

  9. [17]

    P.; and Ba, J

    Kingma, D. P.; and Ba, J. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980

    2014 arXiv

  10. [18]

    Kong, X.; Wang, K.; Wang, S.; Wang, X.; Jiang, X.; Guo, Y.; Shen, G.; Chen, X.; and Ni, Q. 2021. Real-time mask identification for COVID-19 : An edge-computing-based deep learning framework. IEEE Internet of Things Journal, 8(21): 15929--15938

    2021

  11. [19]

    Lai, W.-S.; Huang, J.-B.; Ahuja, N.; and Yang, M.-H. 2017. Deep laplacian pyramid networks for fast and accurate super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 624--632

    2017

  12. [20]

    P.; Castillo, C

    Lau, C. P.; Castillo, C. D.; and Chellappa, R. 2021. Atfacegan: Single face semantic aware image restoration and recognition from atmospheric turbulence. IEEE Transactions on Biometrics, Behavior, and Identity Science, 3(2): 240--251

    2021

  13. [21]

    Li, D.; Xu, C.; Zhang, K.; Yu, X.; Zhong, Y.; Ren, W.; Suominen, H.; and Li, H. 2021. Arvo: Learning all-range volumetric correspondence for video deblurring. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 7721--7731

    2021

  14. [22]

    Li, W.; Tao, X.; Guo, T.; Qi, L.; Lu, J.; and Jia, J. 2020. Mucan: Multi-correspondence aggregation network for video super-resolution. In Proceedings of the European Conference on Computer Vision, 335--351

    2020

  15. [23]

    Li, X.; Liu, M.; Ye, Y.; Zuo, W.; Lin, L.; and Yang, R. 2018. Learning warped guidance for blind face restoration. In Proceedings of the European Conference on Computer Vision, 272--289

    2018

  16. [24]

    Liang, J.; Cao, J.; Fan, Y.; Zhang, K.; Ranjan, R.; Li, Y.; Timofte, R.; and Van Gool, L. 2024. Vrt: A video restoration transformer. IEEE Transactions on Image Processing

    2024

  17. [25]

    Luo, Z.; Yu, L.; Mo, X.; Li, Y.; Jia, L.; Fan, H.; Sun, J.; and Liu, S. 2021. Ebsr: Feature enhanced burst super-resolution with deformable alignment. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 471--478

    2021

  18. [26]

    S.; and Zisserman, A

    Nagrani, A.; Chung, J. S.; and Zisserman, A. 2017. Voxceleb: a large-scale speaker identification dataset. arXiv preprint arXiv:1706.08612

    2017 arXiv

  19. [27]

    S.; Vemulapalli, R.; and Brown, M

    Sajjadi, M. S.; Vemulapalli, R.; and Brown, M. 2018. Frame-recurrent video super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 6626--6634

    2018

  20. [28]

    Tian, Y.; Zhang, Y.; Fu, Y.; and Xu, C. 2020. Tdan: Temporally-deformable alignment network for video super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 3360--3369

    2020

  21. [29]

    C.; Yu, K.; Dong, C.; and Change Loy, C

    Wang, X.; Chan, K. C.; Yu, K.; Dong, C.; and Change Loy, C. 2019. Edvr: Video restoration with enhanced deformable convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, 0--0

    2019

  22. [30]

    Wang, Z.; Zhang, Z.; Zhang, X.; Zheng, H.; Zhou, M.; Zhang, Y.; and Wang, Y. 2023. DR2 : Diffusion-based robust degradation remover for blind face restoration. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 1704--1713

    2023

  23. [31]

    Yang, T.; Ren, P.; Xie, X.; and Zhang, L. 2021. Gan prior embedded network for blind face restoration in the wild. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 672--681

    2021

  24. [32]

    Yu, J.; Zhang, B.; Kuang, Z.; Lin, D.; and Fan, J. 2016. iPrivacy : Image privacy protection by identifying sensitive objects via deep multi-task learning. IEEE Transactions on Information Forensics and Security, 12(5): 1005--1016

    2016

  25. [33]

    M.; and Huang, T

    Zhang, H.; Yang, J.; Zhang, Y.; Nasrabadi, N. M.; and Huang, T. S. 2011. Close the loop: Joint blind image restoration and recognition with sparse representation prior. In International Conference on Computer Vision, 770--777

    2011

  26. [34]

    Zhang, X.; and Wu, X. 2022. Multi-modality deep restoration of extremely compressed face videos. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(2): 2024--2037

    2022

  27. [35]

    Zhang, X.; Wu, X.; Zhai, X.; Ben, X.; and Tu, C. 2020. Davd-net: Deep audio-aided video decompression of talking heads. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 12335--12344

    2020

  28. [36]

    Zhou, K.; Li, W.; Lu, L.; Han, X.; and Lu, J. 2022. Revisiting temporal alignment for video restoration. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 6053--6062

    2022

Pith tools

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