pith. sign in

arxiv: 2604.09989 · v1 · submitted 2026-04-11 · 💻 cs.CV · cs.AI

FlowPalm: Optical Flow Driven Non-Rigid Deformation for Geometrically Diverse Palmprint Generation

Yuchen Zou , Huikai Shao , Lihuang Fang , Zhipeng Xiong , Dexing Zhong This is my paper

Pith reviewed 2026-05-10 16:35 UTC · model grok-4.3

classification 💻 cs.CV cs.AI
keywords palmprint generationoptical flownon-rigid deformationdiffusion modelsynthetic biometric datageometric variationpalmprint recognition
0
0 comments X

The pith

FlowPalm estimates optical flows between real palmprint pairs and applies them progressively in diffusion to generate geometrically diverse synthetic palmprints that improve recognition accuracy.

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

The paper seeks to create synthetic palmprints that include realistic non-rigid geometric changes in addition to style variations. Current generation techniques rely on basic handcrafted changes for geometry, which limits their usefulness for training recognition models. FlowPalm extracts deformation patterns from optical flows computed on actual palmprint pairs and incorporates those patterns through a controlled progressive sampling step inside a diffusion process. The resulting images preserve identity while adding the complex deformations seen in real palms. Experiments across six datasets show that models trained on these images outperform those trained on data from prior generation methods.

Core claim

FlowPalm estimates optical flows between pairs of real palmprints to capture statistical patterns of geometric deformations, then uses a progressive sampling schedule during diffusion that gradually applies these deformations while preserving identity consistency, producing synthetic palmprints that reflect both style and complex non-rigid variation.

What carries the argument

Optical-flow-driven progressive sampling process that injects real deformation statistics into diffusion generation.

If this is right

  • Synthetic palmprint datasets can replace larger collections of real images for training recognition systems.
  • Recognition models gain robustness to pose and deformation changes that simple augmentations miss.
  • Generation frameworks can now target both texture and shape variation in a single pipeline.
  • Downstream tasks on the six tested benchmarks achieve higher accuracy when using FlowPalm data.
  • The method reduces the need to collect additional real palmprints under varied conditions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same optical-flow prior could be adapted to generate training data for other non-rigid biometric traits such as fingerprints or faces under expression changes.
  • By producing identity-consistent yet geometrically varied samples, the approach supports privacy-preserving dataset creation without exposing additional real images.
  • Extending the progressive sampling to video sequences of palms might allow synthesis of dynamic deformation sequences for video-based recognition.

Load-bearing premise

Optical flows computed between existing real palmprint pairs already contain the full range of geometric deformations that occur in real palms and can be introduced without breaking identity.

What would settle it

Train a recognition model on FlowPalm-generated images and test it on a dataset containing palmprints with geometric deformations outside the range of the training pairs; if accuracy does not exceed the baseline using only simple augmentations, the claim is falsified.

Figures

Figures reproduced from arXiv: 2604.09989 by Dexing Zhong, Huikai Shao, Lihuang Fang, Yuchen Zou, Zhipeng Xiong.

Figure 1
Figure 1. Figure 1: Overview of the proposed method and its performance [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the deformation prior construction. Defor [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Statistical distribution and visualization of deformation [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Overview of the proposed deformation-driven three-stage generation strategy. A deformation field sampled from the deformation [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Visualization of the intermediate results of our [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Visualization of the effects of crease and noise warping. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
read the original abstract

Recently, synthetic palmprints have been increasingly used as substitutes for real data to train recognition models. To be effective, such synthetic data must reflect the diversity of real palmprints, including both style variation and geometric variation. However, existing palmprint generation methods mainly focus on style translation, while geometric variation is either ignored or approximated by simple handcrafted augmentations. In this work, we propose FlowPalm, an optical-flow-driven palmprint generation framework capable of simulating the complex non-rigid deformations observed in real palms. Specifically, FlowPalm estimates optical flows between real palmprint pairs to capture the statistical patterns of geometric deformations. Building on these priors, we design a progressive sampling process that gradually introduces the geometric deformations during diffusion while maintaining identity consistency. Extensive experiments on six benchmark datasets demonstrate that FlowPalm significantly outperforms state-of-the-art palmprint generation approaches in downstream recognition tasks. Project page: https://yuchenzou.github.io/FlowPalm/

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript presents FlowPalm, a framework for generating synthetic palmprints that incorporates non-rigid geometric deformations using optical flow estimated from real palmprint pairs. It employs a progressive sampling process in a diffusion model to introduce these deformations while aiming to maintain identity consistency. The authors claim that this approach leads to superior performance in downstream palmprint recognition tasks compared to state-of-the-art methods across six benchmark datasets.

Significance. If the central claims hold, the work addresses an important gap in synthetic biometric data generation by providing a data-driven mechanism for geometric variation beyond style transfer or simple augmentations. This could improve the diversity and utility of synthetic palmprints for training recognition models, with the optical-flow prior offering a principled way to capture real deformation statistics.

major comments (2)
  1. [Method description (progressive sampling)] The progressive sampling process is described as maintaining identity consistency while introducing deformations, but the method provides no explicit identity-preservation loss, no embedding-distance monitoring across diffusion steps, and no verification metrics (e.g., EER or cosine similarity) between source and generated palms. This is load-bearing for the claim that downstream recognition gains arise from modeled non-rigid deformations rather than style variation or artifacts.
  2. [Experiments and results] The abstract asserts outperformance on six benchmark datasets, yet the provided details lack quantitative tables, ablation studies isolating the optical-flow component, baseline implementation specifics, or error analysis. Without these, the support for the central claim of significant improvement cannot be verified.
minor comments (1)
  1. [Abstract] Ensure the full manuscript includes self-contained experimental details rather than relying on the project page for reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. The comments highlight important areas for clarification and expansion, which we address point by point below. We will revise the manuscript accordingly to strengthen the presentation of our method and results.

read point-by-point responses

  1. Referee: [Method description (progressive sampling)] The progressive sampling process is described as maintaining identity consistency while introducing deformations, but the method provides no explicit identity-preservation loss, no embedding-distance monitoring across diffusion steps, and no verification metrics (e.g., EER or cosine similarity) between source and generated palms. This is load-bearing for the claim that downstream recognition gains arise from modeled non-rigid deformations rather than style variation or artifacts.

    Authors: We agree that the current manuscript does not include an explicit identity-preservation loss term or direct monitoring of embedding distances across diffusion steps. Identity consistency is instead maintained implicitly by conditioning the diffusion model on the source palmprint and using a progressive sampling schedule that applies optical flow deformations gradually. This design choice was intended to avoid over-constraining the generation while leveraging the diffusion prior. To address the concern directly, we will add verification metrics in the revised version, including cosine similarity between source and generated palm embeddings as well as EER computed on identity preservation across generated samples. These additions will provide explicit evidence that performance improvements derive from the modeled non-rigid deformations. revision: yes

  2. Referee: [Experiments and results] The abstract asserts outperformance on six benchmark datasets, yet the provided details lack quantitative tables, ablation studies isolating the optical-flow component, baseline implementation specifics, or error analysis. Without these, the support for the central claim of significant improvement cannot be verified.

    Authors: The manuscript reports results across six benchmark datasets with comparisons to state-of-the-art methods, but we acknowledge that the current presentation would benefit from greater detail. In the revision we will expand the experiments section to include complete quantitative tables, dedicated ablation studies that isolate the optical-flow prior (e.g., performance with and without the flow-driven deformation component), explicit baseline implementation details (hyperparameters, training protocols), and error analysis covering variance across datasets and representative failure cases. These additions will make the empirical support fully verifiable. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The described method estimates optical flows directly from real palmprint pairs as input priors and incorporates them into a diffusion-based progressive sampling process. This constitutes a standard data-driven generative pipeline with no equations or claims that reduce to self-definitions, fitted parameters renamed as predictions, or load-bearing self-citations. Identity consistency is asserted as a design goal of the sampling process rather than derived tautologically from prior results. Downstream benchmark evaluations provide independent assessment of the output.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review limited to abstract; no explicit free parameters, invented entities, or additional axioms are stated beyond the core domain assumption that optical flow captures deformation statistics.

axioms (1)
  • domain assumption Optical flows between real palmprint pairs capture the statistical patterns of geometric deformations
    This is the load-bearing premise invoked to justify the progressive sampling process.

pith-pipeline@v0.9.0 · 5475 in / 1088 out tokens · 55060 ms · 2026-05-10T16:35:45.088401+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

50 extracted references · 50 canonical work pages

  1. [1]

    Idiff-face: Synthetic-based face recognition through fizzy identity-conditioned diffusion model

    Fadi Boutros, Jonas Henry Grebe, Arjan Kuijper, and Naser Damer. Idiff-face: Synthetic-based face recognition through fizzy identity-conditioned diffusion model. InProc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), pages 19650– 19661, 2023. 2

    work page 2023

  2. [2]

    Go-with-the-flow: Motion- controllable video diffusion models using real-time warped noise

    Ryan Burgert, Yuancheng Xu, Wenqi Xian, Oliver Pilarski, Pascal Clausen, Mingming He, Li Ma, Yitong Deng, Lingx- iao Li, Mohsen Mousavi, et al. Go-with-the-flow: Motion- controllable video diffusion models using real-time warped noise. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recog- nit. (CVPR), pages 13–23, 2025. 3

    work page 2025

  3. [3]

    Pascal Chang, Jingwei Tang, Markus Gross, and Vinicius C. Azevedo. How I warped your noise: a temporally-correlated noise prior for diffusion models. InProc. Int. Conf. Learn. Represent. (ICLR), 2024. 3, 5

    work page 2024

  4. [4]

    Mo- bilefacenets: Efficient cnns for accurate real-time face veri- fication on mobile devices

    Sheng Chen, Yang Liu, Xiang Gao, and Zhen Han. Mo- bilefacenets: Efficient cnns for accurate real-time face veri- fication on mobile devices. InProc. Chin. Conf. Biometric Recognit. (CCBR), pages 428–438, 2018. 6

    work page 2018

  5. [5]

    Randaugment: Practical automated data augmentation with a reduced search space

    Ekin D Cubuk, Barret Zoph, Jonathon Shlens, and Quoc V Le. Randaugment: Practical automated data augmentation with a reduced search space. InProc. IEEE/CVF Conf. Com- put. Vis. Pattern Recognit. Workshops (CVPRW), pages 702– 703, 2020. 7

    work page 2020

  6. [6]

    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. InProc. IEEE/CVF Conf. Comput. Vis. Pat- tern Recognit. (CVPR), pages 4690–4699, 2019. 2

    work page 2019

  7. [7]

    Diffusion models beat gans on image synthesis.Adv

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis.Adv. Neural Inf. Process. Syst. (NeurIPS), 34:8780–8794, 2021. 3

    work page 2021

  8. [8]

    Amgnet: Aligned multilevel gabor convolu- tion network for palmprint recognition.IEEE Trans

    Dandan Fan, Xu Liang, Chunsheng Zhang, Wei Jia, and David Zhang. Amgnet: Aligned multilevel gabor convolu- tion network for palmprint recognition.IEEE Trans. Circuits Syst. Video Technol., 34:4175–4189, 2023. 1

    work page 2023

  9. [9]

    Generative adversarial networks.Commun

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial networks.Commun. ACM, 63:139–144, 2020. 2

    work page 2020

  10. [10]

    Multi- spectral palm image fusion for accurate contact-free palm- print recognition

    Ying Hao, Zhenan Sun, Tieniu Tan, and Chao Ren. Multi- spectral palm image fusion for accurate contact-free palm- print recognition. InProc. IEEE Int. Conf. Image Process. (ICIP), pages 281–284, 2008. 5

    work page 2008

  11. [11]

    Denoising dif- fusion probabilistic models.Adv

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models.Adv. Neural Inf. Process. Syst. (NeurIPS), 33:6840–6851, 2020. 2, 3

    work page 2020

  12. [12]

    Pce-palm: Palm crease energy based two-stage realis- tic pseudo-palmprint generation

    Jianlong Jin, Lei Shen, Ruixin Zhang, Chenglong Zhao, Ge Jin, Jingyun Zhang, Shouhong Ding, Yang Zhao, and Wei Jia. Pce-palm: Palm crease energy based two-stage realis- tic pseudo-palmprint generation. InProc. AAAI Conf. Artif. Intell. (AAAI), pages 2616–2624, 2024. 1, 2, 5, 6, 7, 8

    work page 2024

  13. [13]

    Diff-palm: Realistic palm- print generation with polynomial creases and intra-class vari- ation controllable diffusion models

    Jianlong Jin, Chenglong Zhao, Ruixin Zhang, Sheng Shang, Jianqing Xu, Jingyun Zhang, ShaoMing Wang, Yang Zhao, Shouhong Ding, Wei Jia, et al. Diff-palm: Realistic palm- print generation with polynomial creases and intra-class vari- ation controllable diffusion models. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pages 26367– 26376, 2025. ...

    work page 2025

  14. [14]

    Unified adversarial augmentation for improving palmprint recognition

    Jianlong Jin, Chenglong Zhao, Ruixin Zhang, Sheng Shang, Yang Zhao, Jun Wang, Jingyun Zhang, Shouhong Ding, Wei Jia, and Yunsheng Wu. Unified adversarial augmentation for improving palmprint recognition. InProc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), pages 14141–14151, 2025. 2, 6, 7

    work page 2025

  15. [15]

    Synthetic data in human analysis: A survey.IEEE Trans

    Indu Joshi, Marcel Grimmer, Christian Rathgeb, Christoph Busch, Francois Bremond, and Antitza Dantcheva. Synthetic data in human analysis: A survey.IEEE Trans. Pattern Anal. Mach. Intell., 46:4957–4976, 2024. 2

    work page 2024

  16. [16]

    Adaface: Quality adaptive margin for face recognition

    Minchul Kim, Anil K Jain, and Xiaoming Liu. Adaface: Quality adaptive margin for face recognition. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pages 18750–18759, 2022. 2

    work page 2022

  17. [17]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A method for stochastic optimization. InProc. Int. Conf. Learn. Represent. (ICLR), 2015. 6

    work page 2015

  18. [18]

    Text- to-image diffusion models are great sketch-photo matchmak- ers

    Subhadeep Koley, Ayan Kumar Bhunia, Aneeshan Sain, Pinaki Nath Chowdhury, Tao Xiang, and Yi-Zhe Song. Text- to-image diffusion models are great sketch-photo matchmak- ers. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pages 16826–16837, 2024. 3

    work page 2024

  19. [19]

    It’s all about your sketch: Democratising sketch control in diffusion models

    Subhadeep Koley, Ayan Kumar Bhunia, Deeptanshu Sekhri, Aneeshan Sain, Pinaki Nath Chowdhury, Tao Xiang, and Yi- Zhe Song. It’s all about your sketch: Democratising sketch control in diffusion models. InProc. IEEE/CVF Conf. Com- put. Vis. Pattern Recognit. (CVPR), pages 7204–7214, 2024. 3

    work page 2024

  20. [20]

    Personal identification us- ing multibiometrics rank-level fusion.IEEE Trans

    Ajay Kumar and Sumit Shekhar. Personal identification us- ing multibiometrics rank-level fusion.IEEE Trans. Syst. Man Cybern., 41:743–752, 2010. 5

    work page 2010

  21. [21]

    Equivdm: Equivariant video diffusion models with temporally consistent noise.arXiv preprint arXiv:2504.09789, 2025

    Chao Liu and Arash Vahdat. Equivdm: Equivariant video diffusion models with temporally consistent noise.arXiv preprint arXiv:2504.09789, 2025. 3

    work page arXiv 2025

  22. [22]

    Data pro- tection in palmprint recognition via dynamic random invisi- ble watermark embedding.IEEE Trans

    Chengcheng Liu, Dexing Zhong, and Huikai Shao. Data pro- tection in palmprint recognition via dynamic random invisi- ble watermark embedding.IEEE Trans. Circuits Syst. Video Technol., 32:6927–6940, 2022. 1

    work page 2022

  23. [23]

    Sf2net: Sequence feature fusion network for palmprint verification.IEEE Trans

    Yunlong Liu, Lu Leng, Ziyuan Yang, Andrew Beng Jin Teoh, and Bob Zhang. Sf2net: Sequence feature fusion network for palmprint verification.IEEE Trans. Inf. Forensics Secur., 20: 9936–9949, 2025. 1

    work page 2025

  24. [24]

    Multidiff: Consistent novel view synthesis 9 from a single image

    Norman M ¨uller, Katja Schwarz, Barbara R ¨ossle, Lorenzo Porzi, Samuel Rota Bulo, Matthias Nießner, and Peter Kontschieder. Multidiff: Consistent novel view synthesis 9 from a single image. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pages 10258–10268, 2024. 3

    work page 2024

  25. [25]

    GLIDE: towards photorealis- tic image generation and editing with text-guided diffusion models

    Alexander Quinn Nichol, Prafulla Dhariwal, Aditya Ramesh, Pranav Shyam, Pamela Mishkin, Bob McGrew, Ilya Sutskever, and Mark Chen. GLIDE: towards photorealis- tic image generation and editing with text-guided diffusion models. InProc. Int. Conf. Mach. Learn. (ICML), pages 16784–16804, 2022. 3

    work page 2022

  26. [26]

    Hierarchical cross-modal image generation for mul- timodal biometric recognition with missing modality.IEEE Trans

    Zaiyu Pan, Shuangtian Jiang, Xiao Yang, Hai Yuan, and Jun Wang. Hierarchical cross-modal image generation for mul- timodal biometric recognition with missing modality.IEEE Trans. Inf. Forensics Secur., 20:4308–4321, 2025. 1

    work page 2025

  27. [27]

    Arc2face: A foundation model for id-consistent human faces

    Foivos Paraperas Papantoniou, Alexandros Lattas, Stylianos Moschoglou, Jiankang Deng, Bernhard Kainz, and Stefanos Zafeiriou. Arc2face: A foundation model for id-consistent human faces. InProc. Eur. Conf. Comput. Vis. (ECCV), pages 241–261, 2024. 2

    work page 2024

  28. [28]

    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. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pages 10684– 10695, 2022. 3

    work page 2022

  29. [29]

    Towards cross-dataset palmprint recognition via joint pixel and feature alignment

    Huikai Shao and Dexing Zhong. Towards cross-dataset palmprint recognition via joint pixel and feature alignment. IEEE Trans. Image Process., 30:3764–3777, 2021. 2

    work page 2021

  30. [30]

    Palmgan for cross-domain palmprint recognition

    Huikai Shao, Dexing Zhong, and Yuhan Li. Palmgan for cross-domain palmprint recognition. InProc. IEEE Int. Conf. Multimedia Expo (ICME), pages 1390–1395, 2019. 2

    work page 2019

  31. [31]

    Deep distilla- tion hashing for unconstrained palmprint recognition.IEEE Trans

    Huikai Shao, Dexing Zhong, and Xuefeng Du. Deep distilla- tion hashing for unconstrained palmprint recognition.IEEE Trans. Instrum. Meas., 70:1–13, 2021. 5

    work page 2021

  32. [32]

    Generating stylized features for single-source cross-dataset palmprint recognition with unseen target dataset.IEEE Trans

    Huikai Shao, Pengxu Li, and Dexing Zhong. Generating stylized features for single-source cross-dataset palmprint recognition with unseen target dataset.IEEE Trans. Image Process., 33:4911–4922, 2024. 1

    work page 2024

  33. [33]

    Rpg-palm: Realistic pseudo-data generation for palmprint recognition

    Lei Shen, Jianlong Jin, Ruixin Zhang, Huaen Li, Kai Zhao, Yingyi Zhang, Jingyun Zhang, Shouhong Ding, Yang Zhao, and Wei Jia. Rpg-palm: Realistic pseudo-data generation for palmprint recognition. InProc. IEEE/CVF Int. Conf. Com- put. Vis. (ICCV), pages 19605–19616, 2023. 1, 2, 6, 7

    work page 2023

  34. [34]

    Denois- ing diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denois- ing diffusion implicit models. InProc. Int. Conf. Learn. Rep- resent. (ICLR), 2021. 4

    work page 2021

  35. [35]

    Raft: Recurrent all-pairs field transforms for optical flow

    Zachary Teed and Jia Deng. Raft: Recurrent all-pairs field transforms for optical flow. InProc. Eur. Conf. Comput. Vis. (ECCV), pages 402–419, 2020. 2, 3, 6

    work page 2020

  36. [36]

    Facea-net: Facial attribute-driven id preserving im- age generation network

    Jiayu Wang, Yue Yu, Jingjing Chen, Qi Dai, and Yu-Gang Jiang. Facea-net: Facial attribute-driven id preserving im- age generation network. InProc. AAAI Conf. Artif. Intell. (AAAI), pages 7736–7743, 2025. 2

    work page 2025

  37. [37]

    Id 3: Identity-preserving- yet-diversified diffusion models for synthetic face recogni- tion

    Jianqing Xu, Shen Li, Jiaying Wu, Miao Xiong, Ailin Deng, Jiazhen Ji, Yuge Huang, Guodong Mu, Wenjie Feng, Shouhong Ding, and Bryan Hooi. Id 3: Identity-preserving- yet-diversified diffusion models for synthetic face recogni- tion. InAdv. Neural Inf. Process. Syst. (NeurIPS), 2024. 2

    work page 2024

  38. [38]

    Consistent flow distillation for text-to-3d generation

    Runjie Yan, Yinbo Chen, and Xiaolong Wang. Consistent flow distillation for text-to-3d generation. InProc. Int. Conf. Learn. Represent. (ICLR), 2025. 3

    work page 2025

  39. [39]

    Physics-driven spectrum-consistent federated learning for palmprint verification.Int

    Ziyuan Yang, Andrew Beng Jin Teoh, Bob Zhang, Lu Leng, and Yi Zhang. Physics-driven spectrum-consistent federated learning for palmprint verification.Int. J. Comput. Vis., 132: 4253–4268, 2024. 1

    work page 2024

  40. [40]

    An online system of multispectral palm- print verification.IEEE Trans

    David Zhang, Zhenhua Guo, Guangming Lu, Lei Zhang, and Wangmeng Zuo. An online system of multispectral palm- print verification.IEEE Trans. Instrum. Meas., 59:480–490,

    work page

  41. [41]

    Towards contactless palmprint recognition: A novel device, a new benchmark, and a collaborative representation based identification approach.Pattern Recognit., 69:199– 212, 2017

    Lin Zhang, Lida Li, Anqi Yang, Ying Shen, and Meng Yang. Towards contactless palmprint recognition: A novel device, a new benchmark, and a collaborative representation based identification approach.Pattern Recognit., 69:199– 212, 2017. 5

    work page 2017

  42. [42]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV), pages 3836–3847, 2023. 3

    work page 2023

  43. [43]

    Pay by showing your palm: A study of palmprint verification on mobile platforms

    Yingyi Zhang, Lin Zhang, Xiao Liu, Shengjie Zhao, Ying Shen, and Yukai Yang. Pay by showing your palm: A study of palmprint verification on mobile platforms. InProc. IEEE Int. Conf. Multimedia Expo (ICME), pages 862–867, 2019. 5

    work page 2019

  44. [44]

    Towards palmprint verification on smartphones.arXiv preprint arXiv:2003.13266, 2020

    Yingyi Zhang, Lin Zhang, Ruixin Zhang, Shaoxin Li, Jilin Li, and Feiyue Huang. Towards palmprint verification on smartphones.arXiv preprint arXiv:2003.13266, 2020. 5

    work page arXiv 2003

  45. [45]

    B´ezierpalm: A free lunch for palmprint recognition

    Kai Zhao, Lei Shen, Yingyi Zhang, Chuhan Zhou, Tao Wang, Ruixin Zhang, Shouhong Ding, Wei Jia, and Wei Shen. B´ezierpalm: A free lunch for palmprint recognition. InProc. Eur. Conf. Comput. Vis. (ECCV), pages 19–36, 2022. 2

    work page 2022

  46. [46]

    Structure suture learning-based robust multiview palmprint recognition.IEEE Trans

    Shuping Zhao, Lunke Fei, Jie Wen, Bob Zhang, Pengyang Zhao, and Shuyi Li. Structure suture learning-based robust multiview palmprint recognition.IEEE Trans. Neural Net- works Learn. Syst., 35:8401–8413, 2022. 1

    work page 2022

  47. [47]

    Regpalm: Toward large-scale open-set palmprint recognition by reducing pattern variance

    Yaoyao Zhong, Weilong Chai, Libin Wang, Dandan Zheng, Huiyuan Fu, and Huadong Ma. Regpalm: Toward large-scale open-set palmprint recognition by reducing pattern variance. IEEE Trans. Inf. Forensics Secur., 20:8541–8554, 2025. 2

    work page 2025

  48. [48]

    Alias-free latent diffusion models: Improving fractional shift equivariance of diffusion latent space

    Yifan Zhou, Zeqi Xiao, Shuai Yang, and Xingang Pan. Alias-free latent diffusion models: Improving fractional shift equivariance of diffusion latent space. InProc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pages 34–44,

    work page

  49. [49]

    Contactless palmprint image recognition across smartphones with self- paced cyclegan.IEEE Trans

    Qi Zhu, Guangnan Xin, Lunke Fei, Dong Liang, Zheng Zhang, Daoqiang Zhang, and David Zhang. Contactless palmprint image recognition across smartphones with self- paced cyclegan.IEEE Trans. Inf. Forensics Secur., 18:4944– 4954, 2023. 2

    work page 2023

  50. [50]

    Pfig-palm: Controllable palmprint generation via pixel and feature identity guidance

    Yuchen Zou, Huikai Shao, Chengcheng Liu, Siyu Zhu, Zongqing Hou, and Dexing Zhong. Pfig-palm: Controllable palmprint generation via pixel and feature identity guidance. IEEE Trans. Image Process., 34:6603–6615, 2025. 1, 2, 6, 7 10

    work page 2025