Pith. sign in

REVIEW 4 major objections 5 minor 50 references

MambaStyle: Efficient StyleGAN Inversion for Real Image Editing with State-Space Models

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read MambaStyle uses vision state-space models to invert real images into StyleGAN's latent space in 0.023 seconds with a 104.56M-parameter encoder, matching the inversion quality of much larger encoders while enabling fast attribute edits.

desk verdict Efficient VSSM-based GAN inversion encoder with real speed and parameter gains; quality claims are overstated relative to its own table. read the letter →

arxiv 2505.15822 v1 pith:RGXTTPUL submitted 2025-05-06 eess.IV cs.CVcs.LG

Jhon Lopez , Carlos Hinojosa , Henry Arguello , Bernard Ghanem This is my paper · ORCID
classification eess.IVcs.CVcs.LG
keywords GANinversionStyleGAN2state-spacemodelsvisionimageeditingencoder-basedMambareal-time
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

Real-image editing through StyleGAN has been held back by a three-way trade-off: optimization-based inversion is accurate but slow, encoder-based inversion is fast but loses detail, and the most detailed encoders are heavy. This paper claims a single-stage encoder built on vision state-space models (VSSMs) breaks that trade-off, reconstructing and editing images while using far fewer parameters and less compute than the top-quality prior encoders. It reports inversion with L2=0.001 and MS-SSIM=0.986 on CelebA-HQ, second-best perceptual scores after a much larger method, and the fastest inference (0.023 s per image) among compared encoders. The same network both inverts and edits, conditioned on an edit direction d, and the whole design is trained on paired images synthesized from pretrained StyleGAN2 noise samples rather than on real photographs.

What carries the argument

The load-bearing object is the Vision State-Space Module (VSSM), built on the 2D selective scan (SS2D) operation: it expands an image patch into four directional sequences, processes each with the input-dependent S6 state-space update from Mamba, and recombines the results, giving linear complexity in sequence length. Three such modules refine the multi-scale features of the encoder, and the Fuser combines those features with the editing direction d using 1D SSM, 2D convolutions, and a VSSM module before passing through modulated convolution blocks to produce the fused feature map that the pretrained StyleGAN2 generator consumes. These two modules together convert the problem of editing a real image into a learned, single-pass mapping from (image, edit direction) to (latent code, feature map).

What would settle it

Retrain the same MambaStyle architecture on synthetic pairs drawn from a different base distribution or with an added distribution shift such as severe compression or out-of-domain faces, then evaluate inversion on the standard real-image test sets; if LPIPS and FID on real photos degrade substantially relative to the reported 7.575 FID on CelebA-HQ, the synthetic-only training assumption is falsified.

Watch

Extended reading notes

Core claim

The paper's central claim is that an encoder built from vision state-space models can map a real image into both a StyleGAN2 style code in W+ space and a spatial feature map at layer k=9, so the pretrained generator reconstructs the image or produces an edited version from the same representation. When the editing direction d is zero the pipeline inverts; when d is nonzero it generates an edited image. The contribution is the architecture: a multi-scale IResNet-based encoder whose hierarchical features are refined by VSSM blocks, plus a Fuser module that injects the edit direction into the feature map via 1D SSM, convolutional, and modulated-convolution layers. With 104.56M total parameters and 157.82 GMACs, the paper reports the best L2 and MS-SSIM inversion results on CelebA-HQ among the compared encoders, second-best LPIPS and FID behind a much heavier method, and faster per-image inference than every compared baseline, with a new state of the art on the car domain.

Load-bearing premise

The load-bearing premise is that paired training images generated entirely from random noise with a pretrained StyleGAN2 are a sufficient stand-in for real photographs, so an encoder that never sees real images during training will generalize to real test photos at inference.

Editorial extensions

If this is right

  • Interactive image editing becomes practical: the full inversion-and-edit pipeline runs in about 0.023 s per image, faster than every compared encoder and orders of magnitude faster than optimization-based PTI (124 s).
  • The same trained model handles both tasks: setting the edit direction to zero reconstructs the image, and setting it to a StyleCLIP, GANSpace, or InterfaceGAN direction edits the attribute, with an explicit loss term enforcing that the latent-code change equals the requested direction.
  • High inversion fidelity no longer requires a heavy encoder: with 104.56M parameters and 157.82 GMACs, MambaStyle matches or beats the inversion quality of encoders two to four times larger, so the accuracy gains of feature-space inversion can be deployed in lower-budget settings.
  • The single-stage training recipe uses only synthetic pairs generated from pretrained StyleGAN2, avoiding the multi-phase training and pretrained-encoder initialization used by predecessors such as SFE and e4e.

Reading between the lines

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

  • The authors do not test temporal consistency; if inversion at 0.023 s is applied to video frames, edited attributes could flicker, so measuring per-frame drift would tell whether real-time video editing is actually achievable.
  • Because training uses only StyleGAN2-generated pairs, the encoder's generalization is bounded by the generator's support; replacing the synthetic source with a different generator or mixing in real images is a direct test of whether the single-stage scheme transfers.
  • The Fuser injects the edit direction directly into spatial feature maps, so the same architecture could plausibly take text or reference-image embeddings as the conditioning input, turning the framework into a general conditioning mechanism rather than a direction-only editor.
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. MambaStyle proposes a single-stage, encoder-based StyleGAN inversion and editing framework that uses vision state-space model (VSSM) modules in both a multi-scale encoder and a novel 'Fuser' module. The encoder outputs a W+ latent code and feature maps; the Fuser combines the editing direction with the encoded features and injects them into StyleGAN2 at layer k=9. Training uses paired synthetic images generated from noise with a pretrained StyleGAN2 generator and editing directions, and the model is evaluated on real CelebA-HQ and Stanford Cars test images. The paper claims a superior balance of inversion quality, editing quality, and computational efficiency, reporting 104.56M total parameters, 157.82 GMACs, and 0.023 s inference time.

Significance. If the efficiency and real-image generalization claims hold, the paper provides a useful demonstration that state-space models can serve as efficient encoders for GAN inversion, potentially enabling more interactive applications. The architecture and the efficiency numbers are a genuine contribution. However, the central claim of 'superior inversion and editing results' is contradicted by the paper's own Table 1: SFE achieves better LPIPS, FID, and two of three editing FID scores, while MambaStyle's wins are concentrated in L2, MS-SSIM, and one edit. Moreover, the model is trained exclusively on synthetic StyleGAN images and then applied to real photos, leaving an unverified domain gap. The paper does not provide code or error bars, which limits reproducibility and statistical assessment, but the efficiency advantage is clearly quantified.

major comments (4)
  1. [Abstract; Section 4.2, Table 1] The claim that MambaStyle achieves 'superior inversion and editing results' is not supported by the paper's own numbers. In Table 1, SFE outperforms MambaStyle on LPIPS (0.019 vs. 0.025), FID (3.535 vs. 7.575), and on two of the three editing metrics (Smile(-): 24.388 vs. 27.149; Old(+): 41.677 vs. 42.819). MambaStyle wins on L2, MS-SSIM, and Glasses(+), but these are pixel-level or structural metrics that can be inflated by feature copying through the Fuser. The paper should reframe its contribution as efficiency at competitive quality, and temper the abstract and conclusion accordingly.
  2. [Section 3.2 and Section 4 (Dataset paragraph)] The model is trained exclusively on synthetic pairs (X, Xe) generated from random noise via pretrained StyleGAN2 (Fig. 2, and the text 'generated from noise vectors z without directly using the original images'), yet all quantitative evaluations are performed on real CelebA-HQ and Stanford Cars images. The encoder never observes a real photograph during training, so the reported real-image inversion and editing metrics rest on an unverified assumption of generalization from the StyleGAN manifold to the real-image distribution. The authors need to provide a control experiment on held-out synthetic test images, add real-image fine-tuning, or analyze the distribution shift (for example, reporting FID between the generated training images and the real test set). Without this, the central real-image claims are not established.
  3. [Section 4.2, Tables 1 and 2] No error bars, confidence intervals, or statistical significance tests are reported for any metric. This is particularly important because MambaStyle trails SFE on several metrics and only marginally leads on others (e.g., L2 0.001 vs. 0.002; Glasses(+) 72.518 vs. 73.098). The reader cannot determine whether the differences between methods are meaningful. The authors should report means and standard deviations over multiple runs, or at least bootstrap confidence intervals, for the main comparisons.
  4. [Section 4.3, W/o Fuser ablation] The 'W/o Fuser' ablation removes the Fuser entirely, which causes a large performance drop; this is expected because the Fuser provides the feature injection at layer k=9. However, the paper does not isolate the effect of the Fuser conditional on the same latent code quality. Since the encoder and Fuser are trained jointly, the ablation conflates the Fuser's contribution with changes in the encoder's training dynamics. A more informative ablation would keep the encoder fixed and add/remove only the Fuser path during training, or evaluate the encoder-only output with a frozen Fuser. This is a secondary point, but it would strengthen the causal claim about the Fuser.
minor comments (5)
  1. [Section 3, paragraph after Eq. (5)] The text says 'limiting its flexibility for editions' — 'editions' should be 'editing' or 'edits'.
  2. [Section 4.1, second paragraph] The sentence 'our method effectively preserves the background of images both in inversion and after editing' is followed by a comparison to HyperStyle, but the grammar is incomplete: 'such as HyperStyle, which often fail' should be 'which often fails' or 'which often fail to preserve them'.
  3. [Section 4, Metrics paragraph] The inference time measurements in Table 1 and Table 3 lack a stated hardware configuration. Please specify the GPU model and any warm-up or batch-size details so that times are comparable across the paper and against the literature.
  4. [Section 3.3, Eq. (8)] The values of the weighting coefficients λ1...λ5 are deferred to the supplementary material, but no supplementary file is present in the submission. Please include these values, or make the supplementary material available, since the reported trade-offs depend on them.
  5. [Figure 2 caption] The caption is extremely long and essentially duplicates the method description in Section 3. Please shorten it to a summary and refer the reader to Section 3 for details.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is a supervised regression on paired StyleGAN-synthesized data with external editing directions and external benchmark metrics.

full rationale

MambaStyle's training and evaluation are not connected by construction. The encoder E and Fuser U are trained by minimizing a weighted combination of reconstruction, perceptual, identity, structure, and editing losses (Eq. 8) on paired samples (X, X_e) generated as X = G(w+) and X_e = G(w+ + d) from random noise z, as stated in Section 3 and Fig. 2. The predicted quantities ŵ and F_k are outputs of learned networks, not re-statements of the training labels; the editing loss Le = ||(ŵ_e − ŵ) − d||1 is a regularizer on latent differences, not a term that bakes in the reported FID, LPIPS, L2, or MS-SSIM scores. At inference the model is applied to held-out real CelebA-HQ and Stanford Cars images, and the reported inversion and editing numbers are computed against those real test distributions using metrics defined independently of the training objective. Editing directions come from external methods (InterfaceGAN, StyleCLIP, GANSpace), and baselines use official or previously reported checkpoints. The only self-citation (ref [27]) appears in related work and is not load-bearing. The synthetic-only training data and the resulting domain gap are a legitimate external-validity concern, but they do not make any prediction equivalent to an input by construction. No fitted parameter is renamed as a prediction, and no uniqueness theorem or ansatz is imported from the authors' prior work. The paper is therefore self-contained in its derivation chain, with circularity score 0.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central method is an empirical architecture; it leans on pretrained generators and loss networks, external editing directions, and several hand-chosen hyperparameters. No new physical entities are introduced.

free parameters (3)
  • Loss weighting coefficients lambda_1 through lambda_5 = Not reported in main text; deferred to supplementary
    Tuned separately for inversion and editing tasks (Section 3.3); these values determine the relative strength of reconstruction, perceptual, identity, structure, and editing losses.
  • StyleGAN feature injection layer k = k = 9
    The Fuser output F_k is added as a skip connection at layer k = 9 of StyleGAN (Section 3). This layer choice is selected by hand and is not ablated for other layer indices.
  • VSSM and Fuser architecture hyperparameters = Not specified
    Number of VSSM blocks, hidden widths, patch sizes, and projection head dimensions are not given; a re-implementation must choose them by hand.
assumptions (5)
  • domain assumption Images synthesized from pretrained StyleGAN2 on FFHQ and LSUN Cars are a sufficient training distribution for inverting real photographs.
    All paired training data are generated from noise (Section 3, Fig. 2), while evaluation is on real CelebA-HQ and Stanford Cars images; the paper never tests or discusses this domain shift.
  • domain assumption The pretrained StyleGAN2 generator and mapping network provide a valid latent space for the target real images.
    The method assumes W+ and F_k spaces can faithfully represent real faces and cars; this is inherited from prior GAN inversion work and not revalidated.
  • domain assumption External editing directions (InterfaceGAN, StyleCLIP, GANSpace) are semantically valid for the target attributes.
    These directions generate the paired training images and are later used for evaluation; any error in them propagates into both training and metrics.
  • standard math The S6 selective scan and SS2D vision scan from the cited Mamba and VMamba papers behave as described.
    The architecture is constructed from these modules without modification; the paper relies on their correctness.
  • domain assumption Pretrained ArcFace and U-Net models provide reliable identity and structure losses.
    The identity and structure loss terms use these pretrained networks; their judgments are treated as ground truth.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MambaStyle: Efficient StyleGAN Inversion for Real Image Editing with State-Space Models." pith.science (2026). https://pith.science/paper/RGXTTPUL

@misc{pith2026250515822,
  author       = {Pith},
  title        = {Pith review of: MambaStyle: Efficient StyleGAN Inversion for Real Image Editing with State-Space Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RGXTTPUL}},
  note         = {Machine review of arXiv:2505.15822}
}
read the original abstract

The task of inverting real images into StyleGAN's latent space to manipulate their attributes has been extensively studied. However, existing GAN inversion methods struggle to balance high reconstruction quality, effective editability, and computational efficiency. In this paper, we introduce MambaStyle, an efficient single-stage encoder-based approach for GAN inversion and editing that leverages vision state-space models (VSSMs) to address these challenges. Specifically, our approach integrates VSSMs within the proposed architecture, enabling high-quality image inversion and flexible editing with significantly fewer parameters and reduced computational complexity compared to state-of-the-art methods. Extensive experiments show that MambaStyle achieves a superior balance among inversion accuracy, editing quality, and computational efficiency. Notably, our method achieves superior inversion and editing results with reduced model complexity and faster inference, making it suitable for real-time applications.

Figures

Figures reproduced from arXiv: 2505.15822 by the authors.

Figure 1
Figure 1. Our method encodes real images into the StyleGAN la [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. During training, we leverage the pretrained StyleGAN2 Generator and Mapping network to generate an image X and its edited version Xe from a random noise vector z and an editing direction d. Then, we learn the latent vectors wˆ ∈ W+ and features Fˆk ∈ Fk using our proposed MambaStyle architecture. This architecture comprises (i) a Multi-scale Mamba-based Encoder that encodes the input image X and produces w ′ and fea… view at source ↗
Figure 3
Figure 3. (Right) Vision State-Space Module (VSSM), and (Left) [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Visual comparison of our proposed method with prior encoder-based approaches in the face domain, based on inversion recon [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Visual comparison of our proposed method with previ [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Visual representation of pose rotation across architectural variations of the proposed MambaStyle network. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Qualitative results highlighting the importance of the [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 29 canonical work pages

  1. [1]

    Im- age2stylegan: How to embed images into the stylegan latent space? In Proceedings of the IEEE/CVF international con- ference on computer vision, pages 4432–4441, 2019

    Rameen Abdal, Yipeng Qin, and Peter Wonka. Im- age2stylegan: How to embed images into the stylegan latent space? In Proceedings of the IEEE/CVF international con- ference on computer vision, pages 4432–4441, 2019. 1, 2

    work page 2019

  2. [2]

    Im- age2stylegan++: How to edit the embedded images? In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 8296–8305, 2020

    Rameen Abdal, Yipeng Qin, and Peter Wonka. Im- age2stylegan++: How to edit the embedded images? In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 8296–8305, 2020. 1, 2

    work page 2020

  3. [3]

    Restyle: A residual-based stylegan encoder via iterative refinement

    Yuval Alaluf, Or Patashnik, and Daniel Cohen-Or. Restyle: A residual-based stylegan encoder via iterative refinement. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6711–6720, 2021. 2, 5, 6

    work page 2021

  4. [4]

    Hyperstyle: Stylegan inversion with hy- pernetworks for real image editing

    Yuval Alaluf, Omer Tov, Ron Mokady, Rinon Gal, and Amit Bermano. Hyperstyle: Stylegan inversion with hy- pernetworks for real image editing. In Proceedings of the IEEE/CVF conference on computer Vision and pattern recognition, pages 18511–18521, 2022. 5, 6

    work page 2022

  5. [5]

    High-fidelity gan inversion with padding space

    Qingyan Bai, Yinghao Xu, Jiapeng Zhu, Weihao Xia, Yu- jiu Yang, and Yujun Shen. High-fidelity gan inversion with padding space. InEuropean Conference on Computer Vision, pages 36–53. Springer, 2022. 2, 5, 6

    work page 2022

  6. [6]

    Triplanenet: An encoder for eg3d inversion

    Ananta R Bhattarai, Matthias Nießner, and Artem Sev- astopolsky. Triplanenet: An encoder for eg3d inversion. In Proceedings of the IEEE/CVF Winter Conference on Appli- cations of Computer Vision, pages 3055–3065, 2024. 2

    work page 2024

  7. [7]

    The devil is in the details: Stylefeatureeditor for detail-rich stylegan inversion and high quality image editing

    Denis Bobkov, Vadim Titov, Aibek Alanov, and Dmitry Vetrov. The devil is in the details: Stylefeatureeditor for detail-rich stylegan inversion and high quality image editing. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 9337–9346, 2024. 1, 2, 5, 6, 8

    work page 2024

  8. [8]

    Transformers are ssms: General- ized models and efficient algorithms through structured state space duality

    Tri Dao and Albert Gu. Transformers are ssms: General- ized models and efficient algorithms through structured state space duality. arXiv preprint arXiv:2405.21060, 2024. 1, 4

    arXiv 2024

Show all 50 references
  1. [9]

    Arcface: Additive angular margin loss for deep face recognition

    Jiankang Deng, Jia Guo, Niannan Xue, and Stefanos Zafeiriou. Arcface: Additive angular margin loss for deep face recognition. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 4690–4699, 2019. 5

    2019

  2. [10]

    Hyperinverter: Improving stylegan inversion via hy- pernetwork

    Tan M Dinh, Anh Tuan Tran, Rang Nguyen, and Binh-Son Hua. Hyperinverter: Improving stylegan inversion via hy- pernetwork. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 11389– 11398, 2022. 2, 5, 6

    2022

  3. [11]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl- vain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition ...

    2021

  4. [12]

    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. Advances in neural information processing systems, 27, 2014. 1

    2014

  5. [13]

    Mamba: Linear-time sequence modeling with selective state spaces

    Albert Gu and Tri Dao. Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752, 2023. 1, 3, 4

    2023 arXiv

  6. [14]

    Hippo: Recurrent memory with optimal polynomial projections

    Albert Gu, Tri Dao, Stefano Ermon, Atri Rudra, and Christo- pher R´e. Hippo: Recurrent memory with optimal polynomial projections. Advances in neural information processing sys- tems, 33:1474–1487, 2020. 1, 2

    2020

  7. [15]

    Efficiently modeling long sequences with structured state spaces

    Albert Gu, Karan Goel, and Christopher R ´e. Efficiently modeling long sequences with structured state spaces. arXiv preprint arXiv:2111.00396, 2021. 1, 2, 3, 4

    2021 arXiv

  8. [16]

    Ganspace: Discovering interpretable gan con- trols

    Erik H ¨ark¨onen, Aaron Hertzmann, Jaakko Lehtinen, and Sylvain Paris. Ganspace: Discovering interpretable gan con- trols. Advances in neural information processing systems , 33:9841–9850, 2020. 5

    2020

  9. [17]

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

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilib- rium. Advances in neural information processing systems , 30, 2017. 5

    2017

  10. [18]

    Style transformer for image inversion and editing

    Xueqi Hu, Qiusheng Huang, Zhengyi Shi, Siyuan Li, Changxin Gao, Li Sun, and Qingli Li. Style transformer for image inversion and editing. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11337–11346, 2022. 1, 2, 5, 6, 7, 8

    2022

  11. [19]

    Progressive growing of gans for improved qual- ity, stability, and variation.arXiv preprint arXiv:1710.10196,

    Tero Karras. Progressive growing of gans for improved qual- ity, stability, and variation.arXiv preprint arXiv:1710.10196,

    arXiv

  12. [20]

    A style-based generator architecture for generative adversarial networks

    Tero Karras, Samuli Laine, and Timo Aila. A style-based generator architecture for generative adversarial networks. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 4401–4410, 2019. 1, 2, 5

    2019

  13. [21]

    Analyzing and improv- ing the image quality of stylegan

    Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Analyzing and improv- ing the image quality of stylegan. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8110–8119, 2020. 5

    2020

  14. [22]

    Alias-free generative adversarial networks

    Tero Karras, Miika Aittala, Samuli Laine, Erik H ¨ark¨onen, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Alias-free generative adversarial networks. Advances in neural infor- mation processing systems, 34:852–863, 2021. 1, 2

    2021

  15. [23]

    3d object representations for fine-grained categorization

    Jonathan Krause, Michael Stark, Jia Deng, and Li Fei-Fei. 3d object representations for fine-grained categorization. In Proceedings of the IEEE international conference on com- puter vision workshops, pages 554–561, 2013. 5

    2013

  16. [24]

    Maskgan: Towards diverse and interactive facial image ma- nipulation

    Cheng-Han Lee, Ziwei Liu, Lingyun Wu, and Ping Luo. Maskgan: Towards diverse and interactive facial image ma- nipulation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5549–5558,

  17. [25]

    Delving style- gan inversion for image editing: A foundation latent space viewpoint

    Hongyu Liu, Yibing Song, and Qifeng Chen. Delving style- gan inversion for image editing: A foundation latent space viewpoint. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 10072– 10082, 2023. 2, 5

    2023

  18. [26]

    Vmamba: Visual state space model

    Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Qixiang Ye, Jianbin Jiao, and Yunfan Liu. Vmamba: Visual state space model. In The Thirty- eighth Annual Conference on Neural Information Processing Systems (NeurIPS), 2024. 3, 4

    2024

  19. [27]

    Privacy-preserving optics for enhancing protection in face de-identification

    Jhon Lopez, Carlos Hinojosa, Henry Arguello, and Bernard Ghanem. Privacy-preserving optics for enhancing protection in face de-identification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 12120–12129, 2024. 2

    2024

  20. [28]

    Styleclip: Text-driven manipulation of stylegan imagery

    Or Patashnik, Zongze Wu, Eli Shechtman, Daniel Cohen-Or, and Dani Lischinski. Styleclip: Text-driven manipulation of stylegan imagery. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 2085–2094,

  21. [29]

    Styleres: Transforming the residuals for real image editing with stylegan

    Hamza Pehlivan, Yusuf Dalva, and Aysegul Dundar. Styleres: Transforming the residuals for real image editing with stylegan. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 1828– 1837, 2023. 1, 2, 5, 6

    2023

  22. [30]

    Adversarial latent autoencoders

    Stanislav Pidhorskyi, Donald A Adjeroh, and Gianfranco Doretto. Adversarial latent autoencoders. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 14104–14113, 2020. 2

    2020

  23. [31]

    Encoding in style: a stylegan encoder for image-to-image translation

    Elad Richardson, Yuval Alaluf, Or Patashnik, Yotam Nitzan, Yaniv Azar, Stav Shapiro, and Daniel Cohen-Or. Encoding in style: a stylegan encoder for image-to-image translation. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 2287–22...

    2021

  24. [32]

    Pivotal tuning for latent-based editing of real im- ages

    Daniel Roich, Ron Mokady, Amit H Bermano, and Daniel Cohen-Or. Pivotal tuning for latent-based editing of real im- ages. ACM Transactions on graphics (TOG) , 42(1):1–13,

  25. [33]

    In- terpreting the latent space of gans for semantic face editing

    Yujun Shen, Jinjin Gu, Xiaoou Tang, and Bolei Zhou. In- terpreting the latent space of gans for semantic face editing. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 9243–9252, 2020. 2, 5

    2020

  26. [34]

    Vmambair: Vi- sual state space model for image restoration

    Yuan Shi, Bin Xia, Xiaoyu Jin, Xing Wang, Tianyu Zhao, Xin Xia, Xuefeng Xiao, and Wenming Yang. Vmambair: Vi- sual state space model for image restoration. arXiv preprint arXiv:2403.11423, 2024. 2, 3

    2024 arXiv

  27. [35]

    Designing an encoder for stylegan image manipulation

    Omer Tov, Yuval Alaluf, Yotam Nitzan, Or Patashnik, and Daniel Cohen-Or. Designing an encoder for stylegan image manipulation. ACM Transactions on Graphics (TOG), 40(4): 1–14, 2021. 1, 2, 4, 5, 6, 8

    2021

  28. [36]

    Mambabyte: Token-free selective state space model

    Junxiong Wang, Tushaar Gangavarapu, Jing Nathan Yan, and Alexander M Rush. Mambabyte: Token-free selective state space model. arXiv preprint arXiv:2401.13660, 2024. 1

    2024 arXiv

  29. [37]

    High-fidelity gan inversion for image attribute editing

    Tengfei Wang, Yong Zhang, Yanbo Fan, Jue Wang, and Qifeng Chen. High-fidelity gan inversion for image attribute editing. In Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition , pages 11379–11388,

  30. [38]

    Mul- tiscale structural similarity for image quality assessment

    Zhou Wang, Eero P Simoncelli, and Alan C Bovik. Mul- tiscale structural similarity for image quality assessment. In The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003, pages 1398–1402. Ieee, 2003. 5

    2003

  31. [39]

    E2style: Improve the efficiency and effectiveness of style- gan inversion

    Tianyi Wei, Dongdong Chen, Wenbo Zhou, Jing Liao, Weiming Zhang, Lu Yuan, Gang Hua, and Nenghai Yu. E2style: Improve the efficiency and effectiveness of style- gan inversion. IEEE Transactions on Image Processing, 31: 3267–3280, 2022. 1, 5

    2022

  32. [40]

    Transformers: State-of-the-art natural language processing

    Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chau- mond, Clement Delangue, Anthony Moi, Pierric Cistac, Tim Rault, R ´emi Louf, Morgan Funtowicz, et al. Transformers: State-of-the-art natural language processing. In Proceedings of the 2020 conference on empirical methods i...

    2020

  33. [41]

    Stylespace analysis: Disentangled controls for stylegan image genera- tion

    Zongze Wu, Dani Lischinski, and Eli Shechtman. Stylespace analysis: Disentangled controls for stylegan image genera- tion. In Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition , pages 12863–12872,

  34. [42]

    Gan inversion: A survey

    Weihao Xia, Yulun Zhang, Yujiu Yang, Jing-Hao Xue, Bolei Zhou, and Ming-Hsuan Yang. Gan inversion: A survey. IEEE transactions on pattern analysis and machine intelli- gence, 45(3):3121–3138, 2022. 1

    2022

  35. [43]

    Segmamba: Long-range sequential modeling mamba for 3d medical image segmentation

    Zhaohu Xing, Tian Ye, Yijun Yang, Guang Liu, and Lei Zhu. Segmamba: Long-range sequential modeling mamba for 3d medical image segmentation. InInternational Conference on Medical Image Computing and Computer-Assisted Interven- tion, pages 578–588. Springer, 2024. 3

    2024

  36. [44]

    Feature-style encoder for style-based gan inversion

    Xu Yao, Alasdair Newson, Yann Gousseau, and Pierre Hel- lier. Feature-style encoder for style-based gan inversion. arXiv preprint arXiv:2202.02183, 2022. 2, 5, 6

    2022 arXiv

  37. [45]

    Lsun: Construction of a large-scale image dataset using deep learning with humans in the loop

    Fisher Yu, Ari Seff, Yinda Zhang, Shuran Song, Thomas Funkhouser, and Jianxiong Xiao. Lsun: Construction of a large-scale image dataset using deep learning with humans in the loop. arXiv preprint arXiv:1506.03365, 2015. 5

    2015 arXiv

  38. [46]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shecht- man, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE conference on computer vision and pattern recogni- tion, pages 586–595, 2018. 5

    2018

  39. [47]

    In- domain gan inversion for real image editing

    Jiapeng Zhu, Yujun Shen, Deli Zhao, and Bolei Zhou. In- domain gan inversion for real image editing. In European conference on computer vision , pages 592–608. Springer,

  40. [48]

    Generative visual manipulation on the natu- ral image manifold

    Jun-Yan Zhu, Philipp Kr ¨ahenb¨uhl, Eli Shechtman, and Alexei A Efros. Generative visual manipulation on the natu- ral image manifold. In Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, Octo- ber 11-14, 2016, Proceedings, Part V 14 , pages 597–...

    2016

  41. [49]

    Vision mamba: Efficient visual representation learning with bidirectional state space model

    Lianghui Zhu, Bencheng Liao, Qian Zhang, Xinlong Wang, Wenyu Liu, and Xinggang Wang. Vision mamba: Efficient visual representation learning with bidirectional state space model. arXiv preprint arXiv:2401.09417, 2024. 3

    2024 arXiv

  42. [50]

    Improved stylegan embedding: Where are the good latents? arXiv preprint arXiv:2012.09036, 2020

    Peihao Zhu, Rameen Abdal, Yipeng Qin, John Femiani, and Peter Wonka. Improved stylegan embedding: Where are the good latents? arXiv preprint arXiv:2012.09036, 2020. 2

    2012 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.