Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Face2QR: A Unified Framework for Aesthetic, Face-Preserving, and Scannable QR Code Generation

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 Face2QR claims that a training-free three-stage pipeline can generate aesthetically styled QR codes that preserve a person's facial identity and scan reliably, with an average face feature distance of 0.51 and scanning success above 94%.

desk verdict Solid engineering paper with a genuinely new IDRS trick, but the SOTA claim and the IDRS feasibility premise need more support than the paper provides. read the letter →

arxiv 2411.19246 v1 pith:GCQGSZXU submitted 2024-11-28 cs.CV

Xuehao Cui , Guangyang Wu , Zhenghao Gan , Guangtao Zhai , Xiaohong Liu This is my paper · ORCID
classification cs.CV
keywords aestheticQRcodesfaceidentitypreservationscannabilityStableDiffusioncontrolnetworksmodulereshufflelatentcodeoptimizationReed-Solomonerrorcorrection
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 a personalized QR code does not have to trade face identity, visual appeal, and scannability against one another. It proposes Face2QR, a three-stage pipeline that generates an initial aesthetic image from a face photo, a text prompt, and a QR message; rebuilds the QR blueprint with the face region frozen; and then refines the latent image with a region-aware loss so the code is reliably decodable. If the claim is right, the practical pay-off is that anyone can produce a decorative, face-recognizable QR image that a phone camera actually scans, with no per-user training. The paper reports an average face feature distance of 0.51, an aesthetic score of 90.1 on AesBench, and scanning success rates above 94% across three decoders and three display sizes. The importance of the claim is that it turns scannability from a strict global constraint into a repairable property of a small set of QR modules.

What carries the argument

The central mechanism is the split of QR modules into protected and malleable sets. IDQR (ID-refined QR integration) uses a Stable-Diffusion-based generator steered by an identity-control network and a QR Controller. IDRS (ID-aware QR ReShuffle) binarizes the generated image, freezes the modules that fall inside the face mask and those belonging to the finder/alignment markers, rearranges the remaining modules so the code matches the target message, and regenerates the Reed-Solomon error correction for the rearranged layout. IDSE (ID-preserved Scannability Enhancement) strengthens the marker patterns and then optimizes the latent code with an adaptive code loss, which applies Gaussian kernels of different widths and different weights to face-region modules versus background modules, together with an aesthetic content loss based on the $L^2$-Wasserstein distance between VGG-19 feature maps of the output and the pre-enhancement image. The entire pipeline is training-free, so the same three stages can be reused with different base generators or identity encoders.

What would settle it

Run the IDRS stage alone on a set of face images and QR payloads with varying lengths, and count how often the constrained shuffle, with the face and marker modules frozen, produces a decodable blueprint before any IDSE refinement. If any realistic face mask leaves too few free modules for the message's error-correction requirement, so that the stage silently raises the error-correction level or must alter the face region, the universality of the pipeline is falsified. Independently, a neutral reproduction of the reported 0.51 average face feature distance using the paper's stated settings would confirm or refute the identity-preservation claim.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is that the three-way conflict between face identity, aesthetics, and QR scannability can be resolved by separating the control signals across three stages instead of asking one model to satisfy all constraints at once. Stage one, IDQR, blends the face into a styled background under two control networks, one for identity and one for the QR pattern. Stage two, IDRS, fixes the QR modules under the face mask and under the finder/alignment markers, rearranges the remaining modules to agree with the target message, and regenerates the error-correction codewords so the new blueprint still decodes. Stage three, IDSE, strengthens the marker patterns and optimizes the latent code with a spatially dynamic loss that is gentle on the face and strict on the background. The authors argue that this division removes the main failure mode of earlier aesthetic-QR methods, which either distorted the face, left the code unscannable, or both.

Load-bearing premise

The load-bearing premise is that, after keeping every black-and-white square of the QR code that overlaps the face or the three corner markers exactly as it is, the remaining squares can always be rearranged, with new error-correction data, to form a valid QR code for the intended message; the paper gives no feasibility analysis, no statement of how much error correction is used, and no failure rate for this step.

Editorial extensions

If this is right

  • If the reported results hold, a single pass through the three stages produces QR images that common phone decoders read successfully in more than 94% of trials at display sizes from 3 cm to 7 cm.
  • Because the pipeline is training-free, new base generators or better identity encoders could be swapped in without retraining the three-stage structure itself.
  • The stage-by-stage error reduction, from 43.85% after stage one to 0.01% after stage three, implies that scannability can be treated as a post-generation repair problem rather than a single end-to-end constraint.
  • The adaptive loss result implies that protecting a semantically important region during latent optimization does not force a global quality-versus-scannability trade-off: different regions can be optimized with different strictness.

Reading between the lines

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

  • A natural extension is to protect any region that must remain visually clean, such as a logo, signature, or product image, by freezing its QR modules and reshuffling only the rest, provided the protected region leaves enough free modules for the required error correction.
  • The paper leaves the feasibility of the IDRS rearrangement unquantified; one could measure, across many face masks and payloads, how often the frozen regions force the regenerated error correction to fail or force a change in the QR error-correction level.
  • If such codes become common, they double as visible biometric tokens, so the same technical success raises a privacy question the paper does not address: the QR image itself now carries a recognizable face in plain sight.
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 / 5 minor

Summary. The paper proposes Face2QR, a three-stage, training-free pipeline for generating aesthetic QR codes that preserve a person's face identity while remaining scannable. Stage 1 (IDQR) uses SDXL with InstantID and a QR Controller to synthesize an initial image. Stage 2 (IDRS) binarizes this image, freezes the modules in the face and marker regions, reshuffles the remaining modules, and regenerates the error-correction codewords to produce a scannable blueprint that guides a second diffusion pass. Stage 3 (IDSE) optimizes the latent code with a spatially adaptive code loss and an aesthetic content loss to enhance scanning robustness while limiting modifications to the face region. Experiments report scan success rates above 94% across three decoders and three display sizes for 20 generated codes, a user study favoring Face2QR, and objective comparisons using ArcFace-based identity distance and AesBench aesthetic scores, leading to a claimed state-of-the-art performance.

Significance. If the central claims hold, Face2QR is a practical contribution: it is a single pipeline that combines face identity, arbitrary aesthetic backgrounds, and scannability, and the paper includes real-world scanning tests with multiple phone decoders, a user study, ablations of the main components, and a code-release URL. The design of the IDRS blueprint and the spatially dynamic loss function are reasonable and clearly motivated by the failure modes of prior work. However, the evidence as presented has several load-bearing gaps—the feasibility of the IDRS constrained optimization is not established, the scanning evaluation uses only 20 hand-picked codes without statistical significance, and the objective error metric may partially mirror the optimization objective—so the SOTA claim is not yet fully supported.

major comments (4)
  1. [§3.1, Eqs. (3)–(4) and §4.1] The feasibility of the IDRS constrained optimization is an unverified premise that the whole pipeline depends on. The constraints freeze the module values in the face and marker regions, which fixes the bits of some data and Reed–Solomon codewords at their physical positions; if those fixed codewords already contain more errors than the ECC can correct, no rearrangement of the remaining modules and no 'regenerated' error correction can make the blueprint decode to the target message. The paper never states the QR ECC level, never reports the fraction of inputs for which a feasible rearranged blueprint exists, and never reports that the blueprint itself decodes before the later stages. Because the scannability claim relies on this step, the authors should add a feasibility analysis over a substantial random sample of face images and messages, report the IDRS failure rate, and verify that the blueprint decodes with a standard decoder.
  2. [§4.3, Table 5] The scannability evidence is based on only 20 hand-picked codes and 50 trials, with no error bars, confidence intervals, or statement of how the codes were selected. The NeurIPS checklist confirms that no statistical significance is reported. As the paper's headline 'average success rate exceeding 94%' and its comparison with Text2QR rely entirely on this table, the authors should report per-code variance, use a larger and more clearly sampled set, and ideally include a significance test against the baseline.
  3. [§4.3, Table 6] The metric d is defined in §4.1 as 'face feature distance' but is computed as a cosine similarity, for which higher values mean better identity preservation; the name is misleading. Moreover, the 0.01 advantage over ArtCoder (0.51 vs. 0.50) is within likely noise, especially because the reported averages are over only 100 samples without error bars. Please clarify the direction and units of d, report standard deviations or confidence intervals, and avoid claiming SOTA on this metric without a significance test.
  4. [§3.2.2, Eqs. (9)–(11) and §4.1] The objective error metric e appears to be computed with the same simulated Gaussian-decoder model that drives the adaptive code loss Lc. If e is computed this way, the quantitative scannability evaluation partly mirrors the optimization objective and may overstate how well the output decodes with real scanners. The authors should specify exactly how e is obtained, and if the simulated decoder is used, add an independent evaluation using the same real decoders as in Table 5 to confirm that the reported error rates transfer to practice.
minor comments (5)
  1. [§3.1 and §4.4] The module is called 'IDRS' in most of the paper but 'IDRE' in these sections; please unify the name throughout.
  2. [§4.3 heading] The heading 'Quantitative Comparsion' should be 'Quantitative Comparison'; also fix the typos 'Scannibility', 'Harmonziation', and 'inherit conflict' elsewhere in the text.
  3. [Table 6] The caption notes '[Key: Best]' but the table does not use boldface or another marker to show which entry is best, which makes the table hard to read.
  4. [§4.1] The definition of e says error modules are counted 'excluding finder and alignment pattern areas,' but the ablation tables also report ef within the face region; please clarify whether e includes modules in the face region or excludes them, since the two statements can be read as contradictory.
  5. [References] Reference [8] is cited as 'Huang et al.' with no initial and an incomplete author list; please provide the full citation for the AesBench tool.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the central scannability and identity-preservation claims are grounded in external real-decoder trials and pretrained face-recognition metrics, with only a mild, non-demonstrable overlap between the internal error count and the IDSE loss.

full rationale

Face2QR's headline claims are supported by independent evidence: real scanning success rates over three smartphone decoders at three display sizes (Table 5), ArcFace-based face-feature similarity d (Table 6), AesBench aesthetic scores, and a 30-participant subjective study (Figure 4). None of these external benchmarks is defined in terms of the method's own optimization targets. The IDRS stage (Eqs. 3-4) does rest on an unverified feasibility premise, namely that freezing face and marker modules still permits a rearrangement of remaining QR modules that decodes to the target message after regenerating error correction; the paper does not report an IDRS failure rate, an ECC level, or direct evidence that the blueprint Ib itself decodes. That is a robustness/completeness gap rather than circularity, because the claimed output is not assumed as an input. The method also relies heavily on Text2QR [43], a prior paper by overlapping authors, for the QR Controller, the blueprint-control idea, and the latent-optimization scheme; however, [43] is a published, externally evaluable system and is cited as a component, not as an unverifiable uniqueness theorem that forces the present result. The only arguable circularity-adjacent point is that the internal error-module metric e (Section 4.1) may be computed with the same simulated Gaussian-kernel decoder that drives the IDSE adaptive code loss (Eqs. 9-11): the paper says 'A simulated decoder [36] using a 2D Gaussian kernel can extract module-wise information consistent with common QR decoders' and then defines the loss with a Gaussian kernel, so the internal e could partially mirror the optimization objective. But the paper never explicitly states that e is measured with that simulated decoder, and the central scannability claim does not rest on e alone—Table 5 reports independent scanning tests with real decoders. Under the rule that circularity must be exhibited by quoting a specific reduction, this overlap is not sufficiently demonstrated to count as a circular step. Overall, the derivation chain is not circular; score 1 reflects the presence of multiple self-citations and an internal metric that may be coupled to the loss, without any load-bearing reduction of the central claim to its own inputs.

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

The central claim rests on feasibility of QR module reshuffling, fidelity of a simulated decoder, behavior of pre-trained components, accuracy of the face mask, and success of latent-space optimization. These are all domain assumptions with no formal guarantees; the paper provides only empirical support. The main hand-tuned parameters are the adaptive loss weights and Gaussian variances, which are ablated in Table 8.

free parameters (5)
  • sigma_f and sigma_b (Gaussian kernel variances in IDSE) = 1.5 and 3.0
    Hand-tuned to balance face distortion and scannability; ablation in Table 8 shows they control face-region artifacts.
  • w_f and w_b (adaptive code loss weights) = 1.0 and 15.0
    Hand-tuned; higher background weight pushes background modules toward the blueprint for scannability.
  • lambda (marker harmonization threshold) = 0.8
    Strengthens finder/alignment patterns; set by default without sensitivity analysis. Eq. 6.
  • Control strengths for InstantID and QR Controller = 0.8 and 1.4
    Control the balance between identity and QR pattern in the SD pipeline; chosen empirically. Section 4.1.
  • Binarization threshold tau = 128
    Standard for 8-bit grayscale, but the blueprint depends on it. Eq. 2.
assumptions (5)
  • domain assumption Reed-Solomon error correction can be regenerated for a QR code with fixed modules (face and markers) to correct any conflicts
    IDRS regenerates error correction codes assuming the reshuffled modules can be made decodable; no feasibility proof is given. Quoted in Section 3.1 around Eqs. 3-4.
  • domain assumption The simulated decoder (2D Gaussian kernel with variance sigma) accurately represents real QR decoders
    Used in IDSE loss (Eq. 9-11) and to count error modules e; inherited from ArtCoder/Text2QR, not re-validated against phone cameras. Section 3.2.2.
  • domain assumption Pre-trained SDXL, InstantID, and QR Controller behave as expected in the training-free pipeline
    The pipeline is training-free and relies on these external models (Eq. 1 and Eq. 5). Section 3.
  • domain assumption The face mask Mf from a pre-trained face recognition model correctly and consistently localizes the face region over the QR grid
    Used to determine fixed modules Delta_f in Eq. 4; if the mask is inaccurate, scannability or identity suffers. Section 3.1.
  • domain assumption Optimizing the latent code in the VQ-VAE latent space preserves aesthetic content while reducing QR error modules
    IDSE assumes a global optimum of Eqs. 7-8 yields both scannable and aesthetic output; only empirical evidence is given. Section 3.2.2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Face2QR: A Unified Framework for Aesthetic, Face-Preserving, and Scannable QR Code Generation." pith.science (2026). https://pith.science/paper/GCQGSZXU

@misc{pith2026241119246,
  author       = {Pith},
  title        = {Pith review of: Face2QR: A Unified Framework for Aesthetic, Face-Preserving, and Scannable QR Code Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GCQGSZXU}},
  note         = {Machine review of arXiv:2411.19246}
}
abstract

Existing methods to generate aesthetic QR codes, such as image and style transfer techniques, tend to compromise either the visual appeal or the scannability of QR codes when they incorporate human face identity. Addressing these imperfections, we present Face2QR-a novel pipeline specifically designed for generating personalized QR codes that harmoniously blend aesthetics, face identity, and scannability. Our pipeline introduces three innovative components. First, the ID-refined QR integration (IDQR) seamlessly intertwines the background styling with face ID, utilizing a unified Stable Diffusion (SD)-based framework with control networks. Second, the ID-aware QR ReShuffle (IDRS) effectively rectifies the conflicts between face IDs and QR patterns, rearranging QR modules to maintain the integrity of facial features without compromising scannability. Lastly, the ID-preserved Scannability Enhancement (IDSE) markedly boosts scanning robustness through latent code optimization, striking a delicate balance between face ID, aesthetic quality and QR functionality. In comprehensive experiments, Face2QR demonstrates remarkable performance, outperforming existing approaches, particularly in preserving facial recognition features within custom QR code designs. Codes are available at $\href{https://github.com/cavosamir/Face2QR}{\text{this URL link}}$.

Figures

Figures reproduced from arXiv: 2411.19246 by the authors.

Figure 1
Figure 1. Face images (first row) and QR code images (second row) generated by Face2QR. Our QR [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The pipeline of Face2QR is a training-free process for generating ID-consistent and [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Illustration of IDRS (left) and IDSE (right). In IDRS, we maintain the information [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: User study of different methods. preservation of face ID. The higher the distance d, the generated face is more consistent with the original face image. We also use AesBench tool [8], which assigns aesthetic scores ranging from 0 to 100 (with higher scores denoting bet…
Figure 5
Figure 5. Figure 5: Curve of different metrics during IDSE. We show metric curves for diverse samples, each [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: The user interface. 2 [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. From Open Loop to Closed Loop: A Test-Time Iterative Optimization Framework for Reference-Consistent Image Generation

    cs.CV 2026-07 conditional novelty 6.5 of 10

    A training-free closed-loop PID controller iteratively corrects latent control signals so diffusion models stay consistent with ID, pose, or depth references better than matched open-loop sampling.

Reference graph

Works this paper leans on

67 extracted references · 52 canonical work pages · cited by 1 Pith paper

  1. [1]

    Wasserstein Generative Adversarial Networks

    Martin Arjovsky, Soumith Chintala, and Léon Bottou. Wasserstein Generative Adversarial Networks. In Proceedings of the 34th International Conference on Machine Learning , 2017

    work page 2017

  2. [2]

    Robust and Unobtrusive Display- to-Camera Communications via Blue Channel Embedding

    Changsheng Chen, Wenjian Huang, Lin Zhang, and Wai Ho Mow. Robust and Unobtrusive Display- to-Camera Communications via Blue Channel Embedding. IEEE Transactions on Image Processing , 28(1):156–169, 2018

    work page 2018

  3. [3]

    PiCode: A New Picture-Embedding 2D Barcode

    Changsheng Chen, Wenjian Huang, Baojian Zhou, Chenchen Liu, and Wai Ho Mow. PiCode: A New Picture-Embedding 2D Barcode. IEEE Transactions on Image Processing, 25(8):3444–3458, 2016

    work page 2016

  4. [4]

    RA Code: A Robust and Aesthetic Code for Resolution-Constrained Applications

    Changsheng Chen, Baojian Zhou, and Wai Ho Mow. RA Code: A Robust and Aesthetic Code for Resolution-Constrained Applications. IEEE Transactions on Circuits and Systems for Video Technology, 28(11):3300–3312, 2018

    work page 2018

  5. [5]

    Halftone QR Codes

    Hung-Kuo Chu, Chia-Sheng Chang, Ruen-Rone Lee, and Niloy J Mitra. Halftone QR Codes. ACM Transactions on Graphics (TOG), 32(6):1–8, 2013

    work page 2013

  6. [6]

    Qartcodes

    Russ Cox. Qartcodes. https://research.swtch.com/qart, 2012

    work page 2012

  7. [7]

    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 conference on computer vision and pattern recognition, pages 4690–4699, 2019

    2019

  8. [8]

    Aesbench: An expert benchmark for multimodal large language models on image aesthetics perception

    Huang et al. Aesbench: An expert benchmark for multimodal large language models on image aesthetics perception. arXiv preprint arXiv: 2401.08276 , 2024

    arXiv 2024

Show all 67 references
  1. [9]

    TERA: Screen-to-Camera Image Code with Transparency, Efficiency, Robustness and Adaptability

    Han Fang, Dongdong Chen, Feng Wang, Zehua Ma, Honggu Liu, Wenbo Zhou, Weiming Zhang, and Neng-Hai Yu. TERA: Screen-to-Camera Image Code with Transparency, Efficiency, Robustness and Adaptability. IEEE Transactions on Multimedia, 24:955–967, 2022

    2022

  2. [10]

    Screen-Shooting Resilient Watermark- ing

    Han Fang, Weiming Zhang, Hang Zhou, Hao Cui, and Nenghai Yu. Screen-Shooting Resilient Watermark- ing. IEEE Transactions on Information F orensics and Security, 14(6):1403–1418, 2018

    2018

  3. [11]

    Stylistic qr code with stable diffusion

    Anthony Fu. Stylistic qr code with stable diffusion. https://antfu.me/posts/ai-qrcode, 2023

    2023

  4. [12]

    The Invisible QR Code

    Zhongpai Gao, Guangtao Zhai, and Chunjia Hu. The Invisible QR Code. In Proceedings of the 23rd ACM International Conference on Multimedia, pages 1047–1050, 2015

    2015

  5. [13]

    QR Images: Optimized Image Embedding in QR Codes

    Gonzalo J Garateguy, Gonzalo R Arce, Daniel L Lau, and Ofelia P Villarreal. QR Images: Optimized Image Embedding in QR Codes. IEEE Transactions on Image Processing, 23(7):2842–2853, 2014

    2014

  6. [14]

    Insightface: A 2d and 3d face analysis project, 2023

    Jia Guo, Xiang An, Jinke Yu, Jing Yang, Alexandros Lattas, Baris Gecer, and Jiankang Deng. Insightface: A 2d and 3d face analysis project, 2023

    2023

  7. [15]

    Lora: Low-rank adaptation of large language models

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. Lora: Low-rank adaptation of large language models. arXiv preprint arXiv:2106.09685, 2021

    2021 arXiv

  8. [16]

    RI- HOOP: Robust Invisible Hyperlinks in Offline and Online Photographs.IEEE Transactions on Cybernetics, pages 1–13, 2020

    Jun Jia, Zhongpai Gao, Kang Chen, Menghan Hu, Xiongkuo Min, Guangtao Zhai, and Xiaokang Yang. RI- HOOP: Robust Invisible Hyperlinks in Offline and Online Photographs.IEEE Transactions on Cybernetics, pages 1–13, 2020

    2020

  9. [17]

    Learning invisible markers for hidden codes in offline-to-online photography

    Jun Jia, Zhongpai Gao, Dandan Zhu, Xiongkuo Min, Guangtao Zhai, and Xiaokang Yang. Learning invisible markers for hidden codes in offline-to-online photography. In Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2022

    2022

  10. [18]

    Analyzing and improving the image quality of StyleGAN

    Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Analyzing and improving the image quality of StyleGAN. In Proc. CVPR, 2020

    2020

  11. [19]

    Adam: A method for stochastic optimization

    Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014

    2014 arXiv

  12. [20]

    Blip-diffusion: Pre-trained subject representation for control- lable text-to-image generation and editing

    Dongxu Li, Junnan Li, and Steven CH Hoi. Blip-diffusion: Pre-trained subject representation for control- lable text-to-image generation and editing. arXiv preprint arXiv:2305.14720, 2023

    2023 arXiv

  13. [21]

    Fastllve: Real-time low-light video enhancement with intensity-aware look-up table

    Wenhao Li, Guangyang Wu, Wenyi Wang, Peiran Ren, and Xiaohong Liu. Fastllve: Real-time low-light video enhancement with intensity-aware look-up table. In ACM Int. Conf. Multimedia, 2023

    2023

  14. [22]

    Photomaker: Customizing realistic human photos via stacked id embedding

    Zhen Li, Mingdeng Cao, Xintao Wang, Zhongang Qi, Ming-Ming Cheng, and Ying Shan. Photomaker: Customizing realistic human photos via stacked id embedding. arXiv preprint arXiv:2312.04461, 2023

    2023 arXiv

  15. [23]

    Oacode: Overall aesthetic 2d barcode on screen

    Zehua Ma, Xi Yang, Han Fang, Weiming Zhang, and Nenghai Yu. Oacode: Overall aesthetic 2d barcode on screen. IEEE Transactions on Multimedia, 2023

    2023

  16. [24]

    Ciagan: Conditional identity anonymization generative adversarial networks

    Maxim Maximov, Ismail Elezi, and Laura Leal-Taixé. Ciagan: Conditional identity anonymization generative adversarial networks. In Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR) , 2020

    2020

  17. [25]

    T2i- adapter: Learning adapters to dig out more controllable ability for text-to-image diffusion models

    Chong Mou, Xintao Wang, Liangbin Xie, Jian Zhang, Zhongang Qi, Ying Shan, and Xiaohu Qie. T2i- adapter: Learning adapters to dig out more controllable ability for text-to-image diffusion models. arXiv preprint arXiv:2302.08453, 2023

    2023 arXiv

  18. [26]

    Glide: Towards photorealistic image generation and editing with text-guided diffusion models

    Alex Nichol, Prafulla Dhariwal, Aditya Ramesh, Pranav Shyam, Pamela Mishkin, Bob McGrew, Ilya Sutskever, and Mark Chen. Glide: Towards photorealistic image generation and editing with text-guided diffusion models. arXiv preprint arXiv:2112.10741, 2021

    2021 arXiv

  19. [27]

    Pexels: Free stock photos, royalty free stock images & videos

    Pexels. Pexels: Free stock photos, royalty free stock images & videos. https://www.pexels.com/, 2024. 11

    2024

  20. [28]

    Hierarchical text-conditional image generation with clip latents

    Aditya Ramesh, Prafulla Dhariwal, Alex Nichol, Casey Chu, and Mark Chen. Hierarchical text-conditional image generation with clip latents. arXiv preprint arXiv:2204.06125, 1(2):3, 2022

    2022 arXiv

  21. [29]

    Polynomial codes over certain finite fields

    Irving S Reed and Gustave Solomon. Polynomial codes over certain finite fields. Journal of the society for industrial and applied mathematics , 8(2):300–304, 1960

    1960

  22. [30]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. High-resolution image synthesis with latent diffusion models. In Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR) , 2022

    2022

  23. [31]

    Davidson, and Jiying Zhao

    Zhihao Shi, Xiaohong Liu, Chengqi Li, Linhui Dai, Jun Chen, Timothy N. Davidson, and Jiying Zhao. Learning for unconstrained space-time video super-resolution. IEEE Trans. Broadcast., 68(2):345–358, 2022

    2022

  24. [32]

    Video frame interpolation via generalized deformable convolution

    Zhihao Shi, Xiaohong Liu, Kangdi Shi, Linhui Dai, and Jun Chen. Video frame interpolation via generalized deformable convolution. IEEE Trans. Multim., 24:426–439, 2022

    2022

  25. [33]

    Video frame interpolation transformer

    Zhihao Shi, Xiangyu Xu, Xiaohong Liu, Jun Chen, and Ming-Hsuan Yang. Video frame interpolation transformer. In Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR) , 2022

    2022

  26. [34]

    Very deep convolutional networks for large-scale image recogni- tion

    Karen Simonyan and Andrew Zisserman. Very deep convolutional networks for large-scale image recogni- tion. In Yoshua Bengio and Yann LeCun, editors, ICCV, 2015

    2015

  27. [35]

    Q-Art Code: Generating Scanning-robust Art-style QR Codes by Deformable Convolution

    Hao Su, Jianwei Niu, Xuefeng Liu, Qingfeng Li, Ji Wan, and Mingliang Xu. Q-Art Code: Generating Scanning-robust Art-style QR Codes by Deformable Convolution. In ACM Int. Conf. Multimedia, 2021

    2021

  28. [36]

    Artcoder: an end-to-end method for generating scanning-robust stylized qr codes

    Hao Su, Jianwei Niu, Xuefeng Liu, Qingfeng Li, Ji Wan, Mingliang Xu, and Tao Ren. Artcoder: an end-to-end method for generating scanning-robust stylized qr codes. In Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2021

    2021

  29. [37]

    Stegastamp: Invisible Hyperlinks in Physical Photographs

    Matthew Tancik, Ben Mildenhall, and Ren Ng. Stegastamp: Invisible Hyperlinks in Physical Photographs. In Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR) , 2020

    2020

  30. [38]

    Face0: Instantaneously conditioning a text-to-image model on a face

    Dani Valevski, Danny Lumen, Yossi Matias, and Yaniv Leviathan. Face0: Instantaneously conditioning a text-to-image model on a face. In SIGGRAPH Asia 2023 Conference Papers, pages 1–10, 2023

    2023

  31. [39]

    Neural discrete representation learning

    Aaron Van Den Oord, Oriol Vinyals, et al. Neural discrete representation learning. Advances in neural information processing systems, 30, 2017

    2017

  32. [40]

    Instantid: Zero-shot identity-preserving generation in seconds

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

    2024 arXiv

  33. [41]

    Robust prior-based single image super resolution under multiple gaussian degradations

    Wenyi Wang, Guangyang Wu, Weitong Cai, Liaoyuan Zeng, and Jianwen Chen. Robust prior-based single image super resolution under multiple gaussian degradations. IEEE Access, 8:74195–74204, 2020

    2020

  34. [42]

    Light Field Messaging with Deep Photographic Steganography

    Eric Wengrowski and Kristin Dana. Light Field Messaging with Deep Photographic Steganography. In Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR) , 2019

    2019

  35. [43]

    Text2qr: Harmonizing aesthetic customization and scanning robustness for text-guided qr code generation

    Guangyang Wu, Xiaohong Liu, Jun Jia, Xuehao Cui, and Guangtao Zhai. Text2qr: Harmonizing aesthetic customization and scanning robustness for text-guided qr code generation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8456–8465, 2024

    2024

  36. [44]

    Accflow: Backward accumulation for long-range optical flow

    Guangyang Wu, Xiaohong Liu, Kunming Luo, Xi Liu, Qingqing Zheng, Shuaicheng Liu, Xinyang Jiang, Guangtao Zhai, and Wenyi Wang. Accflow: Backward accumulation for long-range optical flow. In Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV) , 2023

    2023

  37. [45]

    Pred: A parallel network for handling multiple degradations via single model in single image super-resolution

    Guangyang Wu, Lili Zhao, Wenyi Wang, Liaoyuan Zeng, and Jianwen Chen. Pred: A parallel network for handling multiple degradations via single model in single image super-resolution. In Proc. IEEE Int. Conf. Image Process. (ICIP), 2019

    2019

  38. [46]

    ART-UP: A novel method for generating scanning-robust aesthetic QR codes

    Mingliang Xu, Qingfeng Li, Jianwei Niu, Hao Su, Xiting Liu, Weiwei Xu, Pei Lv, Bing Zhou, and Yi Yang. ART-UP: A novel method for generating scanning-robust aesthetic QR codes. ACM Trans. Multim. Comput. Commun. Appl., 17(1):25:1–25:23, 2021

    2021

  39. [47]

    Stylized aesthetic QR code

    Mingliang Xu, Hao Su, Yafei Li, Xi Li, Jing Liao, Jianwei Niu, Pei Lv, and Bing Zhou. Stylized aesthetic QR code. IEEE Trans. Multim., 21(8):1960–1970, 2019

    1960

  40. [48]

    Prompt-Free Diffusion: Taking" Text" out of Text-to-Image Diffusion Models

    Xingqian Xu, Jiayi Guo, Zhangyang Wang, Gao Huang, Irfan Essa, and Humphrey Shi. Prompt-Free Diffusion: Taking" Text" out of Text-to-Image Diffusion Models. arXiv preprint arXiv:2305.16223, 2023

    2023 arXiv

  41. [49]

    Facestudio: Put your face everywhere in seconds

    Yuxuan Yan, Chi Zhang, Rui Wang, Yichao Zhou, Gege Zhang, Pei Cheng, Gang Yu, and Bin Fu. Facestudio: Put your face everywhere in seconds. arXiv preprint arXiv:2312.02663, 2023

    2023 arXiv

  42. [50]

    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, 2023

    2023 arXiv

  43. [51]

    A3gan: Attribute- aware anonymization networks for face de-identification

    Liming Zhai, Qing Guo, Xiaofei Xie, Lei Ma, Yi Estelle Wang, and Yang Liu. A3gan: Attribute- aware anonymization networks for face de-identification. In Proceedings of the 30th ACM International Conference on Multimedia, pages 5303–5313, 2022

    2022

  44. [52]

    A male man

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 3836–3847, 2023. 12 A Appendix A.1 Additional Experiments A.1.1 Additional Results Our ...

    2023

  45. [53]

    Furthermore, in the last paragraph of the introduction, these contributions are clearly listed again

    Claims Question: Do the main claims made in the abstract and introduction accurately reflect the paper’s contributions and scope? Answer: [Yes] Justification: In abstract, the main contributions of this paper are emphasized. Furthermore, in the last paragraph of the introducti...

  46. [54]

    Limitations

    Limitations Question: Does the paper discuss the limitations of the work performed by the authors? Answer: [Yes] Justification: Please refer to Section 5. Guidelines: • The answer NA means that the paper has no limitation while the answer No means that the paper has limitation...

  47. [55]

    Guidelines: • The answer NA means that the paper does not include theoretical results

    Theory Assumptions and Proofs Question: For each theoretical result, does the paper provide the full set of assumptions and a complete (and correct) proof? Answer: [NA] 6 Justification: This paper does not include theoretical results. Guidelines: • The answer NA means that the...

  48. [56]

    Our code will be released upon publication

    Experimental Result Reproducibility Question: Does the paper fully disclose all the information needed to reproduce the main ex- perimental results of the paper to the extent that it affects the main claims and/or conclusions of the paper (regardless of whether the code and da...

  49. [57]

    Guidelines: • The answer NA means that paper does not include experiments requiring code

    Open access to data and code Question: Does the paper provide open access to the data and code, with sufficient instruc- tions to faithfully reproduce the main experimental results, as described in supplemental material? 7 Answer: [Yes] Justification: We consider publishing th...

  50. [58]

    Guidelines: • The answer NA means that the paper does not include experiments

    Experimental Setting/Details Question: Does the paper specify all the training and test details (e.g., data splits, hyper- parameters, how they were chosen, type of optimizer, etc.) necessary to understand the results? Answer: [Yes] Justification: All experiments details are i...

  51. [59]

    Therefore, the corresponding error bars are not applicable

    Experiment Statistical Significance Question: Does the paper report error bars suitably and correctly defined or other appropriate information about the statistical significance of the experiments? Answer: [No] Justification: This paper mainly conducts qualitative comparisons ...

  52. [60]

    Guidelines: • The answer NA means that the paper does not include experiments

    Experiments Compute Resources Question: For each experiment, does the paper provide sufficient information on the com- puter resources (type of compute workers, memory, time of execution) needed to reproduce the experiments? Answer: [Yes] Justification: Computational resources...

  53. [61]

    Guidelines: • The answer NA means that the authors have not reviewed the NeurIPS Code of Ethics

    Code Of Ethics Question: Does the research conducted in the paper conform, in every respect, with the NeurIPS Code of Ethics https://neurips.cc/public/EthicsGuidelines? Answer: [Yes] Justification: This work is conducted in accordance with the NeurIPS Code of Ethics. Guideline...

  54. [62]

    Guidelines: • The answer NA means that there is no societal impact of the work performed

    Broader Impacts Question: Does the paper discuss both potential positive societal impacts and negative societal impacts of the work performed? Answer: [Yes] Justification: Please refer to the Section 5. Guidelines: • The answer NA means that there is no societal impact of the ...

  55. [63]

    Guidelines: • The answer NA means that the paper poses no such risks

    Safeguards Question: Does the paper describe safeguards that have been put in place for responsible release of data or models that have a high risk for misuse (e.g., pretrained language models, image generators, or scraped datasets)? Answer: [NA] Justification: The paper poses...

  56. [64]

    Guidelines: • The answer NA means that the paper does not use existing assets

    Licenses for existing assets Question: Are the creators or original owners of assets (e.g., code, data, models), used in the paper, properly credited and are the license and terms of use explicitly mentioned and properly respected? Answer: [Yes] Justification: The creators or ...

  57. [65]

    Guidelines: • The answer NA means that the paper does not release new assets

    New Assets Question: Are new assets introduced in the paper well documented and is the documentation provided alongside the assets? Answer: [Yes] Justification: The new assets introduced in the paper are well documented alongside the assets. Guidelines: • The answer NA means t...

  58. [66]

    Guidelines: • The answer NA means that the paper does not involve crowdsourcing nor research with human subjects

    Crowdsourcing and Research with Human Subjects Question: For crowdsourcing experiments and research with human subjects, does the paper include the full text of instructions given to participants and screenshots, if applicable, as well as details about compensation (if any)? A...

  59. [67]

    We have obtained the IRB approval and also adhere to the NeurIPS Code of Ethics

    Institutional Review Board (IRB) Approvals or Equivalent for Research with Human Subjects Question: Does the paper describe potential risks incurred by study participants, whether such risks were disclosed to the subjects, and whether Institutional Review Board (IRB) approvals...

Pith tools

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