Pith. sign in

REVIEW 4 major objections 5 minor 2 cited by

HiFiVFS: High Fidelity Video Face Swapping

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper claims a video face swapping framework, built on Stable Video Diffusion, that preserves fine-grained target attributes and temporal stability through identity-disentangled attribute learning and detailed identity tokens.

desk verdict A plausible diffusion-based video face swapping system whose headline temporal stability claim rests partly on an undefined inference-time 'temporal co-denoising' step that needs to be specified and ablated. read the letter →

arxiv 2411.18293 v2 pith:EYZ7OVIU submitted 2024-11-27 cs.CV

Xu Chen , Keke He , Junwei Zhu , Yanhao Ge , Wei Li , Chengjie Wang This is my paper · ORCID
classification cs.CV
keywords videofaceswappingdiffusionmodelsStabletemporalstabilityfine-grainedattributesidentitydisentanglementtokenizervideo-to-videogeneration
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 is trying to establish that video face swapping can be done directly on video rather than frame-by-frame, using the temporal prior of a pretrained video diffusion model. It claims that previous diffusion-based approaches, which rely on inpainting and global attribute features, lose fine-grained details such as lighting and makeup, and that this loss can be prevented by an identity-desensitization cycle that trains the attribute encoder to ignore who the face is. It also claims that identity similarity improves when the model uses high-resolution intermediate features from a face recognition network instead of the final compact identity vector. If these claims hold, video face swapping gains both temporal stability and the ability to preserve subtle target attributes, which are the two qualities that have limited earlier methods. A human evaluation and quantitative metrics on standard benchmarks are reported as evidence.

What carries the argument

The load-bearing mechanism is the FAL identity-desensitization cycle: the attribute encoder $E_{\text{attr}}$ extracts attribute features $f_{\text{attr}}$ from the target video; the decoder fuses $f_{\text{attr}}$ with a randomly chosen identity vector to synthesize a modified video with a different identity; and $E_{\text{attr}}$ is applied again to that modified video to get $f'_{\text{attr}}$, with the loss $L_{\text{attr}} = \frac{1}{2}\|f_{\text{attr}} - f'_{\text{attr}}\|_2^2$ forcing the two to match. This trains the attribute features to be invariant to identity, and a triplet margin identity loss plus an adversarial loss push the cycle to produce realistic, identity-changed outputs. The second mechanism is DIL, which replaces the usual global 512-dimensional identity vector with 49 spatial tokens taken from the last Res-Block of a face recognition network, fed into cross- and temporal attention to retain fine identity detail.

What would settle it

Take two source faces that are nearly identical in shape but differ in a fine attribute such as makeup or a skin mark, and use them in the FAL cycle: if the swapped output preserves the source's fine attribute instead of the target's, or if the attribute features $f_{\text{attr}}$ and $f'_{\text{attr}}$ differ for two targets with identical attributes but different identities, the identity-disentanglement claim fails.

Watch

Extended reading notes

Core claim

The central discovery is that a video-to-video face swapping pipeline can be built on Stable Video Diffusion by extending it to multi-frame input and conditioning it on identity features, while a Fine-grained Attributes Learning (FAL) module learns attribute features that are disentangled from identity through a reconstruction cycle: the same encoder is applied to the target video and to a synthesized video with a different identity, and the two attribute feature sets are forced to match. This, together with a Detailed Identity Learning (DIL) module that feeds spatial tokens from an intermediate recognition-layer into cross- and temporal attention, is what the authors claim enables both state-of-the-art identity similarity and preservation of fine details such as lighting, makeup, and occlusion handling. In their experiments, HiFiVFS outperforms GAN-based and diffusion-based baselines on FaceForensics++ and on a newly selected VFHQ-FS test set across identity, expression, pose, shape, gaze, video identity distance, and Fréchet Video Distance.

Load-bearing premise

The method assumes that forcing the attribute extractor to output the same features for a video and for the same video with a different face substituted in truly removes the person's identity from those features, rather than the model finding a shortcut where identity still leaks through.

Editorial extensions

If this is right

  • Video face swapping can be treated as a video-to-video generation task, so temporal attention across frames replaces frame-by-frame processing and removes the per-frame jitter that image-based methods suffer.
  • Fine-grained target attributes like lighting, makeup, and occlusion are preserved by injecting low-level attribute features directly into the denoising U-Net input rather than through cross-attention.
  • Identity similarity in swapped videos improves when identity is represented by spatial tokens from an intermediate face-recognition layer instead of the final compact identity vector.
  • The EDM-style denoiser of Stable Video Diffusion lets identity losses be computed on decoded images during training, so the identity and attribute losses integrate directly into the diffusion objective.
  • The method inherits the sampling speed and memory costs of video diffusion models, and the authors point to diffusion distillation as the path to faster synthesis.

Reading between the lines

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

  • The identity-desensitization cycle is a general recipe: any video-to-video task that needs identity-free attribute features (reenactment, expression transfer, makeup transfer) could adopt the same reconstruction cycle with a different identity vector.
  • A stronger check of disentanglement than the reported metrics would be an explicit invariance test: measure how much the attribute features change when the identity is swapped while attributes are held fixed, across many pairs of identities.
  • The 16-frame training clips and sliding temporal co-denosing for long videos suggest a windowed extension to arbitrary-length footage, though boundary effects between windows are not analyzed in the paper.
  • Applying the FAL cycle in latent space rather than pixel space is a deliberate design choice; the paper's own ablation shows that the pixel-space variant drops across all metrics, implying the latent-space alignment with the denoising U-Net is doing real work.
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

4 major / 5 minor

Summary. The paper presents HiFiVFS, a video face swapping framework built on Stable Video Diffusion (SVD). A source image supplies identity and a target video supplies attributes; the model is trained with an inpainting-style data flow, a fine-grained attribute learning module (FAL) that uses an identity-replacement cycle plus adversarial learning, and a detailed identity learning module (DIL) that injects tokens from the last Res-Block layer of a face-recognition model. Training losses include denoising score matching, attribute-cycle, triplet identity, reconstruction, adversarial, and identity-cosine losses. The method is evaluated on FF++ and a new VFHQ-FS test set against GAN- and diffusion-based baselines, using identity retrieval/similarity, expression/shape/gaze/pose errors, VIDD, FVD, qualitative comparisons, and a user study. The paper claims state-of-the-art results in both fidelity and temporal stability.

Significance. If the claims hold, HiFiVFS is a useful advance: it adapts SVD to video-to-video face swapping, introduces a principled attribute-disentanglement objective, and achieves strong quantitative results on external benchmarks. The manuscript is commendable for giving explicit loss definitions, using external datasets and metrics, reporting ablations of FAL and DIL, and providing supplementary video comparisons. The main weakness is that the inference-time 'temporal co-denoising' procedure, which is central to the temporal-stability claim, is not defined or ablated, so the attribution of the reported VIDD/FVD gains to the architectural contributions is currently unverifiable.

major comments (4)
  1. [Sec. 5.1] The term 'temporal co-denoising' appears exactly once: 'the temporal co-denoising is used to weaken the detail discrepancies between different video clips.' It is never defined, no inputs or algorithm are given, and no ablation removes it. Because the model is trained on 16-frame clips but evaluated on 64-frame VFHQ-FS videos, multi-clip inference must be specified; if co-denoising is a test-time clip-chaining or post-processing step, it could be responsible for the VIDD/FVD improvements in Table 2 rather than the proposed FAL, DIL, or SVD temporal attention. Please provide a precise description and an ablation with co-denoising disabled.
  2. [Sec. 4.2, Eq. (3)] The FAL cycle loss Lattr = ||fattr - f'attr||^2 assumes that replacing the identity in Vt with a different identity and then re-extracting attributes yields identity-disentangled attribute features. This assumption is not tested: Eattr could store identity in low-level features that survive the cycle, or Dec could ignore fattr entirely, making the equality hold without true disentanglement. Since fine-grained attribute preservation and identity-leakage prevention are central contributions, the authors should provide a direct validation, for example a face-recognition/ID-retrieval test on fattr alone or an experiment showing that fattr does not recover the source identity.
  3. [Sec. 5.2, Tables 1-3] All headline numbers are point estimates without confidence intervals, repeated-run variance, or significance tests. The user study in Table 3 uses only 15 participants, yet the text states that the model 'significantly outperforms' the baselines, without reporting any statistical test. In addition, the DiffSwap row in Table 1 is based on 'official code' results that 'differ significantly from those in the paper,' as the table caption acknowledges; the authors should state explicitly which set of numbers is used and justify the choice, since a baseline evaluated under a different configuration may not be directly comparable.
  4. [Sec. 1 and Sec. 2.2] The claim that this is 'the first attempt to improve temporal stability within the face swapping framework' (Sec. 1) is contradicted by the paper's own reference [42], MobileFaceSwap, which is explicitly a video face swapping framework. The claim should be narrowed, for example to 'the first diffusion-based video face swapping framework with temporal attention,' or the authors should explain why prior video-based face swapping methods do not count as attempts at temporal stability.
minor comments (5)
  1. [Sec. 4.4] The text says the overall loss has four components and refers to 'Denoising Score Matching (eq. 3),' but Eq. (3) is Lattr; the DSM loss is Eq. (2), and Eq. (8) actually lists three components (LDM, LFAL, Lid).
  2. [Sec. 4.2] The description of FAL says 'we randomly select a face' to obtain frid, but it is not stated whether this face comes from the same video Vt, another video, or a fixed identity bank; please clarify the sampling procedure.
  3. [Sec. 5.2, Human Evaluation] The text mentions 'four reshuffled face swapping results generated by SimSwap, HifiFace, InfoSwap, Face-Adapter, and our HiFiVFS,' which actually lists five methods; rephrase to avoid confusion.
  4. [Sec. 5.1, Evaluation Metrics] VIDD is defined only by a citation to FOS; because VIDD and FVD are central to the temporal-stability claims, the exact computation formula for VIDD should be given in the text or appendix.
  5. [Sec. 5.1, Implementation Details] The phrase 'classifier guidance scale of 2' is unclear because SVD is a latent video diffusion model and no classifier is described; please specify what is being guided and how the guidance scale is applied.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the central SOTA claim rests on external benchmarks and metrics, though the undefined temporal co-denoising is a separate reproducibility concern.

full rationale

The paper's central claim of SOTA video face swapping is supported by external benchmarks (FF++, VFHQ-FS) and external metrics (CosFace for identity, 3DMM coefficients for expression/shape/gaze/pose, VIDD from FOS, FVD from StyleGAN-V), none of which is used as a training loss or fitted target. The FAL cycle in Sec. 4.2, especially Lattr in Eq. (3), is a training objective that enforces attribute-feature invariance under identity replacement; it is an assumption about how to learn disentanglement, not a quantity derived from the same objective and then reported as a prediction. DIL's identity loss in Eq. (7) uses a decoded output and a pre-trained recognition feature, and no fitted parameter is renamed as a prediction. The 3DMM evaluation tool is cited from FaceAdapter, a same-group paper, but it is applied uniformly to all methods and is not used to fit HiFiVFS, so this is a minor self-citation rather than load-bearing evidence. The under-specified 'temporal co-denoising' in Sec. 5.1 is a genuine reproducibility and contribution-attribution risk: the model trains on 16-frame clips and is evaluated on 64-frame VFHQ-FS videos, so this unspecified mechanism could conceivably contribute to VIDD/FVD gains. However, the paper provides no definition or ablation of this step, so there is no quotable reduction of the temporal-stability claim to a fitted parameter or self-citation; under the strict circularity standard, this is an incompleteness concern, not a demonstrated circularity. Self-citations to HifiFace and FaceAdapter as baselines are normal and do not make the derivation circular.

Assumptions & free parameters 10 free parameters · 6 assumptions · 0 invented entities

The central claim rests on pretrained models (SVD, CosFace, BiSeNet, FaceAdapter 3DMM) whose validity is taken from prior work, plus three paper-specific assumptions: the cycle loss truly disentangles identity, SVD temporal attention transfers to face swapping, and the unexplained temporal co-denoising works. Loss weights and inference settings are hand-tuned and not fitted to test metrics.

free parameters (10)
  • lambda_attr = 10
    Weight on Lattr in Eq 6; hand-set.
  • lambda_rec = 10
    Weight on Lrec in Eq 6.
  • lambda_tid = 1
    Weight on Ltid in Eq 6.
  • triplet_margin_m = 0.4
    Margin in Eq 5 for identity triplet loss.
  • lambda_id = 0.1
    Weight on Lid in Eq 8.
  • classifier_guidance_scale = 2
    Sampling guidance scale for 25-step EDM sampler.
  • sampling_steps = 25
    Deterministic EDM sampler steps at inference.
  • train_steps = 55,000 (5,000 warm-up + 50,000)
    Optimization schedule; warm-up fixes Mf to zero.
  • clip_length = 16 frames
    Randomly extracted clip length for training.
  • crop_size = 640x640
    Face-aligned crop size containing more background.
assumptions (6)
  • domain assumption SVD pretrained weights provide a strong temporal and generative prior for face swapping.
    Sec 4.1 initializes UNet from SVD [3]; the temporal stability claim depends on this transfer.
  • ad hoc to paper The FAL cycle loss Lattr enforces identity disentanglement of attribute features.
    Sec 4.2: equality of fattr and f'_attr is assumed to imply identity-free attributes; no direct verification of disentanglement is provided.
  • domain assumption Face parsing (BiSeNet) provides accurate face masks for inpainting.
    Sec 4.1/Fig 3 uses parsed masks to define the inpainting region; failures under occlusion affect fidelity.
  • domain assumption The FaceAdapter 3DMM reconstruction gives unbiased attribute/pose/shape/gaze coefficients for all methods.
    Sec 5.1 uses it for Exp/Shape/Gaze/Pose metrics for all methods.
  • domain assumption CosFace reliably measures identity similarity of swapped faces.
    Sec 5.1 uses CosFace for IDr/IDs.
  • ad hoc to paper Temporal co-denoising maintains consistency across clips longer than the 16-frame training clip.
    Sec 5.1 mentions this procedure but does not define it; long-video stability rests on it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HiFiVFS: High Fidelity Video Face Swapping." pith.science (2026). https://pith.science/paper/EYZ7OVIU

@misc{pith2026241118293,
  author       = {Pith},
  title        = {Pith review of: HiFiVFS: High Fidelity Video Face Swapping},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EYZ7OVIU}},
  note         = {Machine review of arXiv:2411.18293}
}
read the original abstract

Face swapping aims to generate results that combine the identity from the source with attributes from the target. Existing methods primarily focus on image-based face swapping. When processing videos, each frame is handled independently, making it difficult to ensure temporal stability. From a model perspective, face swapping is gradually shifting from generative adversarial networks (GANs) to diffusion models (DMs), as DMs have been shown to possess stronger generative capabilities. Current diffusion-based approaches often employ inpainting techniques, which struggle to preserve fine-grained attributes like lighting and makeup. To address these challenges, we propose a high fidelity video face swapping (HiFiVFS) framework, which leverages the strong generative capability and temporal prior of Stable Video Diffusion (SVD). We build a fine-grained attribute module to extract identity-disentangled and fine-grained attribute features through identity desensitization and adversarial learning. Additionally, We introduce detailed identity injection to further enhance identity similarity. Extensive experiments demonstrate that our method achieves state-of-the-art (SOTA) in video face swapping, both qualitatively and quantitatively.

Figures

Figures reproduced from arXiv: 2411.18293 by the authors.

Figure 1
Figure 1. Face swapping results of HiFiVFS. The face in the source image (orange) is taken to replace the face in the target video (blue). Abstract Face swapping aims to generate results that combine the identity from the source with attributes from the target. Existing methods primarily focus on image-based face swapping. When processing videos, each frame is han￾dled independently, making it difficult to ensure tempo￾ral st… view at source ↗
Figure 2
Figure 2. Training pipeline of face swapping methods. (a) GAN-based methods achieve feature disentanglement by using attribute and [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Pipeline of our proposed HiFiVFS, including training and inference phases. HiFiVFS is primarily trained based on the SVD [3] framework, utilizing multi-frame input and a temporal attention to ensure the stability of the generated videos. In the training phase, HiFiVFS introduces fine-grained attribute learning (FAL) and detailed identity learning (DIL). In FAL, attribute disentanglement and enhancement are achieved … view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: VFHQ-FS results compared with other methods. The source image of each example is placed in the corresponding top-left [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: FF++ results compared with FaceShifter [ [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Face swapping results on wild face videos under various challenging conditions. Our method is capable of producing results with [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Detail Network Structures of FAL 1 [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: More comparisons on FF++. 2 [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 9
Figure 9. Figure 9: More comparisons on FF++. 3 [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Adaptive Identity Anchoring: Closed-Loop Keyframe Placement for Synthetic Paired Supervision in Video Face Swapping

    cs.CV 2026-07 conditional novelty 6.0 of 10

    For video face swapping, adaptively adding swapped anchor frames at the moments of worst identity drift should make synthetic training pairs more faithful than the current first-and-last-frame-only scheme.

  2. CanonSwap: High-Fidelity and Consistent Video Face Swapping via Canonical Space Modulation

    cs.CV 2025-07 conditional novelty 5.0 of 10

    A face swapping framework that decouples motion from appearance by performing identity transfer in a canonical space, improving temporal consistency and identity preservation.

Reference graph

Works this paper leans on

51 extracted references · 40 canonical work pages · cited by 2 Pith papers

  1. [42]

    Mobilefaceswap: A lightweight framework for video face swapping

    Zhiliang Xu, Zhibin Hong, Changxing Ding, Zhen Zhu, Junyu Han, Jingtuo Liu, and Errui Ding. Mobilefaceswap: A lightweight framework for video face swapping. In Pro- ceedings of the AAAI Conference on Artificial Intelligence ,

  2. [1]

    Creating a photoreal digital actor: The digital emily project

    Oleg Alexander, Mike Rogers, William Lambeth, Matt Chi- ang, and Paul Debevec. Creating a photoreal digital actor: The digital emily project. In 2009 Conference for Visual Me- dia Production. IEEE, 2009. 1

    work page 2009

  3. [2]

    A morphable model for the synthesis of 3d faces

    V olker Blanz and Thomas Vetter. A morphable model for the synthesis of 3d faces. In Proceedings of the 26th an- nual conference on Computer graphics and interactive tech- niques, 1999. 3

    1999

  4. [3]

    Stable video diffusion: Scaling latent video diffusion models to large datasets

    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, 3, 4

    arXiv 2023

  5. [4]

    Align your latents: High-resolution video synthesis with la- tent diffusion models

    Andreas Blattmann, Robin Rombach, Huan Ling, Tim Dock- horn, Seung Wook Kim, Sanja Fidler, and Karsten Kreis. Align your latents: High-resolution video synthesis with la- tent diffusion models. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, 2023. 3

    work page 2023

  6. [5]

    Simswap: An efficient framework for high fidelity face swapping

    Renwang Chen, Xuanhong Chen, Bingbing Ni, and Yanhao Ge. Simswap: An efficient framework for high fidelity face swapping. In Proceedings of the 28th ACM International Conference on Multimedia, 2020. 1, 2, 3, 6, 7

    work page 2020

  7. [6]

    Towards real-world video face restoration: A new benchmark

    Ziyan Chen, Jingwen He, Xinqi Lin, Yu Qiao, and Chao Dong. Towards real-world video face restoration: A new benchmark. arXiv preprint arXiv:2404.19500, 2024. 6

    arXiv 2024

  8. [7]

    Stargan v2: Diverse image synthesis for multiple domains

    Yunjey Choi, Youngjung Uh, Jaejun Yoo, and Jung-Woo Ha. Stargan v2: Diverse image synthesis for multiple domains. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2020. 5

    work page 2020

Show all 51 references
  1. [8]

    V oxceleb2: Deep speaker recognition

    Joon Son Chung, Arsha Nagrani, and Andrew Zisserman. V oxceleb2: Deep speaker recognition. arXiv preprint arXiv:1806.05622, 2018. 5

    2018 arXiv

  2. [9]

    Retinaface: Single-shot multi- level face localisation in the wild

    Jiankang Deng, Jia Guo, Evangelos Ververas, Irene Kot- sia, and Stefanos Zafeiriou. Retinaface: Single-shot multi- level face localisation in the wild. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020. 6

    2020

  3. [10]

    Accurate 3d face reconstruction with weakly-supervised learning: From single image to image set

    Yu Deng, Jiaolong Yang, Sicheng Xu, Dong Chen, Yunde Jia, and Xin Tong. Accurate 3d face reconstruction with weakly-supervised learning: From single image to image set. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2019. 2, 3

    2019

  4. [11]

    Diffusion models beat gans on image synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. Advances in neural informa- tion processing systems, 2021. 3

    2021

  5. [12]

    Information bottleneck disentanglement for identity swapping

    Gege Gao, Huaibo Huang, Chaoyou Fu, Zhaoyang Li, and Ran He. Information bottleneck disentanglement for identity swapping. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021. 1, 2, 3, 6, 7

    2021

  6. [13]

    Generative adversarial nets

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In Advances in neural information processing systems, 2014. 2

    2014

  7. [14]

    Face adapter for pre-trained diffusion models with fine-grained id and attribute control

    Yue Han, Junwei Zhu, Keke He, Xu Chen, Yanhao Ge, Wei Li, Xiangtai Li, Jiangning Zhang, Chengjie Wang, and Yong Liu. Face adapter for pre-trained diffusion models with fine-grained id and attribute control. arXiv preprint arXiv:2405.12970, 2024. 1, 2, 3, 4, 6, 7

    2024 arXiv

  8. [15]

    Image-to-image translation with conditional adver- sarial networks

    Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, and Alexei A Efros. Image-to-image translation with conditional adver- sarial networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017. 2

    2017

  9. [16]

    Elucidating the design space of diffusion-based generative models

    Tero Karras, Miika Aittala, Timo Aila, and Samuli Laine. Elucidating the design space of diffusion-based generative models. Advances in neural information processing systems,

  10. [17]

    Smooth- swap: A simple enhancement for face-swapping with smoothness

    Jiseob Kim, Jihoon Lee, and Byoung-Tak Zhang. Smooth- swap: A simple enhancement for face-swapping with smoothness. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022. 1

    2022

  11. [18]

    Faceinpainter: High fidelity face adaptation to heteroge- neous domains

    Jia Li, Zhaoyang Li, Jie Cao, Xingguang Song, and Ran He. Faceinpainter: High fidelity face adaptation to heteroge- neous domains. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021

    2021

  12. [19]

    Faceshifter: Towards high fidelity and occlusion aware face swapping

    Lingzhi Li, Jianmin Bao, Hao Yang, Dong Chen, and Fang Wen. Faceshifter: Towards high fidelity and occlusion aware face swapping. arXiv preprint arXiv:1912.13457, 2019. 1, 2, 3, 6, 7

    1912 arXiv

  13. [20]

    Towards a si- multaneous and granular identity-expression control in per- sonalized face generation

    Renshuai Liu, Bowen Ma, Wei Zhang, Zhipeng Hu, Changjie Fan, Tangjie Lv, Yu Ding, and Xuan Cheng. Towards a si- multaneous and granular identity-expression control in per- sonalized face generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition ,

  14. [21]

    Fine-grained face swapping via regional gan inversion

    Zhian Liu, Maomao Li, Yong Zhang, Cairong Wang, Qi Zhang, Jue Wang, and Yongwei Nie. Fine-grained face swapping via regional gan inversion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023. 1

    2023

  15. [22]

    Styleface: Towards identity-disentangled face generation on megapixels

    Yuchen Luo, Junwei Zhu, Keke He, Wenqing Chu, Ying Tai, Chengjie Wang, and Junchi Yan. Styleface: Towards identity-disentangled face generation on megapixels. In Eu- ropean Conference on Computer Vision. Springer, 2022. 2

    2022

  16. [23]

    Fsgan: Subject agnostic face swapping and reenactment

    Yuval Nirkin, Yosi Keller, and Tal Hassner. Fsgan: Subject agnostic face swapping and reenactment. In ICCV, 2019. 2

    2019

  17. [24]

    Portraitbooth: A versatile portrait model for fast identity-preserved personalization

    Xu Peng, Junwei Zhu, Boyuan Jiang, Ying Tai, Donghao Luo, Jiangning Zhang, Wei Lin, Taisong Jin, Chengjie Wang, and Rongrong Ji. Portraitbooth: A versatile portrait model for fast identity-preserved personalization. InProceedings of the IEEE/CVF Conference on Computer Vision a...

    2024

  18. [25]

    Learn- ing transferable visual models from natural language super- vision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learn- ing transferable visual models from natural language super- vision. In International conference on machine learning...

    2021

  19. [26]

    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. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022. 3, 4, 5 9

    2022

  20. [27]

    Facedancer: Pose-and occlusion-aware high fidelity face swapping

    Felix Rosberg, Eren Erdal Aksoy, Fernando Alonso- Fernandez, and Cristofer Englund. Facedancer: Pose-and occlusion-aware high fidelity face swapping. In Proceedings of the IEEE/CVF winter conference on applications of com- puter vision, 2023. 2

    2023

  21. [28]

    Visual cryptography for bio- metric privacy

    Arun Ross and Asem Othman. Visual cryptography for bio- metric privacy. IEEE transactions on information forensics and security, 2010. 1

    2010

  22. [29]

    Faceforen- sics++: Learning to detect manipulated facial images

    Andreas Rossler, Davide Cozzolino, Luisa Verdoliva, Chris- tian Riess, Justus Thies, and Matthias Nießner. Faceforen- sics++: Learning to detect manipulated facial images. In Proceedings of the IEEE International Conference on Com- puter Vision, 2019. 2, 5

    2019

  23. [30]

    Blendface: Re-designing identity encoders for face- swapping

    Kaede Shiohara, Xingchao Yang, and Takafumi Take- tomi. Blendface: Re-designing identity encoders for face- swapping. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023. 1, 2, 3, 6, 7

    2023

  24. [31]

    Few-shot head swapping in the wild

    Changyong Shu, Hemao Wu, Hang Zhou, Jiaming Liu, Zhibin Hong, Changxing Ding, Junyu Han, Jingtuo Liu, Er- rui Ding, and Jingdong Wang. Few-shot head swapping in the wild. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022. 1

    2022

  25. [32]

    Stylegan-v: A continuous video generator with the price, image quality and perks of stylegan2

    Ivan Skorokhodov, Sergey Tulyakov, and Mohamed Elho- seiny. Stylegan-v: A continuous video generator with the price, image quality and perks of stylegan2. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022. 6

    2022

  26. [33]

    Face2face: Real-time face capture and reenactment of rgb videos

    Justus Thies, Michael Zollhofer, Marc Stamminger, Chris- tian Theobalt, and Matthias Nießner. Face2face: Real-time face capture and reenactment of rgb videos. In Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016. 3

    2016

  27. [34]

    Cosface: Large margin cosine loss for deep face recognition

    Hao Wang, Yitong Wang, Zheng Zhou, Xing Ji, Dihong Gong, Jingchao Zhou, Zhifeng Li, and Wei Liu. Cosface: Large margin cosine loss for deep face recognition. In Pro- ceedings of the IEEE conference on computer vision and pat- tern recognition, 2018. 6

    2018

  28. [35]

    Instantid: Zero-shot identity-preserving generation in seconds

    Qixun Wang, Xu Bai, Haofan Wang, Zekui Qin, Anthony Chen, Huaxia Li, Xu Tang, and Yao Hu. Instantid: Zero-shot identity-preserving generation in seconds. arXiv preprint arXiv:2401.07519, 2024. 3

    2024 arXiv

  29. [36]

    Hififace: 3d shape and semantic prior guided high fidelity face swapping

    Yuhan Wang, Xu Chen, Junwei Zhu, Wenqing Chu, Ying Tai, Chengjie Wang, Jilin Li, Yongjian Wu, Feiyue Huang, and Rongrong Ji. Hififace: 3d shape and semantic prior guided high fidelity face swapping. arXiv preprint arXiv:2106.09965, 2021. 1, 2, 3, 6, 7

    2021 arXiv

  30. [37]

    Fastcomposer: Tuning-free multi- subject image generation with localized attention

    Guangxuan Xiao, Tianwei Yin, William T Freeman, Fr ´edo Durand, and Song Han. Fastcomposer: Tuning-free multi- subject image generation with localized attention. Interna- tional Journal of Computer Vision, 2024. 3

    2024

  31. [38]

    Vfhq: A high-quality dataset and bench- mark for video face super-resolution

    Liangbin Xie, Xintao Wang, Honglun Zhang, Chao Dong, and Ying Shan. Vfhq: A high-quality dataset and bench- mark for video face super-resolution. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022. 5, 6

    2022

  32. [39]

    Designing one unified framework for high-fidelity face reenactment and swapping

    Chao Xu, Jiangning Zhang, Yue Han, Guanzhong Tian, Xi- anfang Zeng, Ying Tai, Yabiao Wang, Chengjie Wang, and Yong Liu. Designing one unified framework for high-fidelity face reenactment and swapping. In European conference on computer vision. Springer, 2022. 1

    2022

  33. [40]

    Region-aware face swapping

    Chao Xu, Jiangning Zhang, Miao Hua, Qian He, Zili Yi, and Yong Liu. Region-aware face swapping. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 7632–7641, 2022

    2022

  34. [41]

    High-resolution face swapping via latent semantics disentanglement

    Yangyang Xu, Bailin Deng, Junle Wang, Yanqing Jing, Jia Pan, and Shengfeng He. High-resolution face swapping via latent semantics disentanglement. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022

    2022

  35. [43]

    Styleswap: Style-based generator empow- ers robust face swapping

    Zhiliang Xu, Hang Zhou, Zhibin Hong, Ziwei Liu, Jiaming Liu, Zhizhi Guo, Junyu Han, Jingtuo Liu, Errui Ding, and Jingdong Wang. Styleswap: Style-based generator empow- ers robust face swapping. In European Conference on Com- puter Vision. Springer, 2022. 2

    2022

  36. [44]

    Ip- adapter: Text compatible image prompt adapter for text-to- image diffusion models

    Hu Ye, Jun Zhang, Sibo Liu, Xiao Han, and Wei Yang. Ip- adapter: Text compatible image prompt adapter for text-to- image diffusion models. arXiv preprint arXiv:2308.06721,

    arXiv

  37. [45]

    Bisenet: Bilateral segmentation network for real-time semantic segmentation

    Changqian Yu, Jingbo Wang, Chao Peng, Changxin Gao, Gang Yu, and Nong Sang. Bisenet: Bilateral segmentation network for real-time semantic segmentation. In Proceed- ings of the European conference on computer vision (ECCV),

  38. [46]

    Celebv-text: A large-scale facial text-video dataset

    Jianhui Yu, Hao Zhu, Liming Jiang, Chen Change Loy, Wei- dong Cai, and Wayne Wu. Celebv-text: A large-scale facial text-video dataset. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, 2023. 5

    2023

  39. [47]

    Diffswap: High-fidelity and con- trollable face swapping via 3d-aware masked diffusion

    Wenliang Zhao, Yongming Rao, Weikang Shi, Zuyan Liu, Jie Zhou, and Jiwen Lu. Diffswap: High-fidelity and con- trollable face swapping via 3d-aware masked diffusion. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, 2023. 1, 2, 3, 4, 6

    2023

  40. [48]

    One shot face swapping on megapixels

    Yuhao Zhu, Qi Li, Jian Wang, Cheng-Zhong Xu, and Zhenan Sun. One shot face swapping on megapixels. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021. 2 10 HiFiVFS: High Fidelity Video Face Swapping Supplementary Material

    2021

  41. [49]

    Network Structures of FAL The detailed structure of our HiFiVFS is shown in Fig. 7. The encoder Eattr consists of three layers, each contain- ing two residual blocks and two self-attention mechanisms. The output fattr is obtained from the last layer of Eattr, while flow repres...

  42. [50]

    More Results For FF++, we put more comparations in Fig. 8 and 9. For VFHQ-FS and wild cases, we have included additional comparative results in the zip file (Comparisons VHFQ-FS and Comparisons Wild). Besides the academic methods mentioned in the main text, we also conducted a...

  43. [51]

    Broader Impact and Limitations Broader Impact. HiFiVFS is capable of consistently pro- ducing high-quality face-swapping videos, even in highly challenging situations, which expands the potential appli- cations of face-swapping technology. However, the risk of misuse poses ser...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.