Ridgeformer: Mutli-Stage Contrastive Training For Fine-grained Cross-Domain Fingerprint Recognition

Bhavin Jawade; Shubham Pandey; Srirangaraj Setlur

REVIEW 3 major objections 6 minor 35 references

Ridgeformer: Mutli-Stage Contrastive Training For Fine-grained Cross-Domain Fingerprint Recognition

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Ridgeformer claims best-reported contactless-to-contact fingerprint verification on HKPolyU, with EER 2.83%.

desk verdict A solid two-stage ViT fingerprint matcher with a real but fixable overclaim: their own Table 1 contradicts the 'outperforms all' headline on TAR@FAR=0.01. read the letter →

arxiv 2506.01806 v1 pith:3OXS2IAZ submitted 2025-06-02 cs.CV cs.AI

Shubham Pandey , Bhavin Jawade , Srirangaraj Setlur This is my paper

classification cs.CVcs.AI

keywords contactlessfingerprintrecognitioncross-domainmatchingvisiontransformercross-attentionmulti-similaritylossbiometricverificationidentificationRidgeformer

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

Ridgeformer sets out to solve cross-domain fingerprint matching, where a fingerprint photographed by a contactless sensor must be matched against a contact-based scan of the same finger. The authors propose a two-stage transformer pipeline: a Vision Transformer extracts global patch-level features, and a cross-attention stage aligns those patches between a probe and a gallery image before computing similarity. Training uses a multi-similarity loss applied to contactless-to-contactless, contactless-to-contact, and contact-to-contact similarity matrices. On HKPolyU contactless-to-contact verification, the method reports an equal error rate of 2.83% and a true-accept rate of 89.34% at FAR=0.01, which the authors state outperforms all previously benchmarked models on that dataset. The claim matters because reliable contactless-to-contact matching is the obstacle to hygienic, touch-free fingerprint authentication in real deployments.

What carries the argument

The machine is a two-stage pipeline named Ridgeformer. Stage 1 uses a Vision Transformer to split each fingerprint into $T$ patch tokens, pools them with global average pooling, and projects the pooled vector through an MLP to get a global embedding. Stage 2 takes the token sets of a probe and a gallery image, concatenates them, and passes them through a cross-attention transformer; the attended tokens are pooled and compared by cosine similarity. Training uses the multi-similarity loss with hard positive/negative mining, computing the loss on three similarity matrices: contactless-to-contactless, contactless-to-contact, and contact-to-contact. The named identity is the combination of the multi-stage architecture and the three-way contrastive objective.

What would settle it

Re-run the HKPolyU contactless-to-contact protocol with all hyperparameters (including $\alpha_{\text{pos}}=2.0$, $\alpha_{\text{neg}}=40.0$, margin, $\tau$, and the learning-rate schedule) chosen only on a held-out validation fold, never on the test split. If the resulting EER rises above 3% or fails to beat MANet's 4.13%, the paper's superiority claim is refuted.

Watch

Extended reading notes

Core claim

The central claim is that a hierarchical global-to-local transformer, trained with a contrastive multi-similarity loss across both same-domain and cross-domain pairs, yields the best reported contactless-to-contact fingerprint verification on HKPolyU, with EER 2.83% and TAR 89.34% at FAR=0.01, beating MANet (EER 4.13%) and a commercial COTS matcher (EER 19.31%). On the RidgeBase benchmark, the same model achieves EER 5.25% for contactless-to-contact and 7.60% for contactless-to-contactless verification, with perfect Rank-1 identification (100%) in the contactless-to-contactless setting. The authors attribute the improvement to Stage 2's cross-attention refinement: the ablation shows EER dropping from 3.74% with only Stage 1 to 2.83% after adding Stage 2 and fine-tuning.

Load-bearing premise

The reported superiority rests on hyperparameters that the authors say were 'determined through experimentation,' with no statement that this tuning was done on held-out validation data rather than the evaluation splits.

Editorial extensions

If this is right

If the reported numbers hold, contactless-to-contact verification is no longer the weak link: a touchless capture can be matched to legacy contact-scan databases at an EER below 3% on HKPolyU.
The ablation shows each stage contributes: global features alone give EER 3.74%, adding cross-attention lowers it to 3.04%, and fine-tuning on the target dataset brings it to 2.83% with TAR rising from 84.16% to 89.34%.
On RidgeBase, the same model reaches 100% Rank-1 identification for contactless-to-contactless matching, suggesting the learned embedding separates identities cleanly in that protocol.
The method's advantage over COTS (EER 19.31% vs 2.83% on HKPolyU) implies that learned cross-domain embeddings can replace or supplement commercial matchers in contactless enrollment scenarios.

Reading between the lines

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

A natural extension, not tested in the paper, is whether the same global-then-cross-attention recipe transfers to other cross-sensor biometrics (e.g., iris or face) where domain shift comes from sensor geometry and illumination rather than fingerprint ridge distortion.
Because the paper does not report validation-based hyperparameter selection, the fairest reading is that the 2.83% EER is an upper bound on what a practitioner would get with a clean tuning protocol; an independent replication with hyperparameters chosen on a disjoint validation split would settle this.
The near-perfect contactless-to-contactless identification on RidgeBase could also be a ceiling effect of the small evaluation set (25 subjects); testing on a larger cohort would reveal whether the 100% Rank-1 figure is meaningful or an artifact of gallery size.
The paper's decision to omit ISPFDv2 results due to protocol non-reproducibility suggests that standardized evaluation protocols for contactless fingerprint benchmarks are a prerequisite for comparing methods; a shared protocol with fixed train/test splits would make future claims directly comparable.

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.

Desk Editor's Note

read the letter

Short version: this is worth reading if you work on contactless fingerprints. The two-stage design—global ViT embedding, then cross-attention between token sets for fine-grained alignment—is sensible, and the ablation shows each stage buys a real EER/TAR improvement. They also ship a code link and are upfront about why ISPFDv2 is not reported (non-reproducible random splits). That is honest.

The headline claim does not survive contact with Table 1. On HKPolyU, Ridgeformer gets EER 2.83% versus ML Fusion's 4.07%, but ML Fusion gets TAR@FAR=0.01 of 94.40% versus Ridgeformer's 89.34%. The abstract and §4.4 say 'outperforms all previously benchmarked models.' That is true for EER, false for the security-relevant operating point they report in the same table. ROC curves cross; either a threshold-preference argument or a qualified 'lowest EER' is needed. This is fixable in revision but it is a real overstatement.

Two other soft spots. First, the paper cites Grosz et al. [15] (minutiae-guided ViT embeddings) and MRA-GNN [16] but does not compare to either; that makes the 'first ViT for contactless-to-contact matching' claim shaky if [15] already does cross-domain work. Second, the hyperparameters are 'determined through experimentation' (§4.1) with no explicit statement that tuning was done on held-out validation, and there are no error bars. Those are standard concerns for this venue, not fatal.

What holds up: the method itself is plausible, the ablation is clean, the Ridgebase numbers look strong, and the code release is a plus. The central architecture is not a gimmick.

Recommendation: send to peer review, but insist the authors reconcile the TAR/EER conflict and either add the cited baselines or justify their omission. As is, the paper should be accepted only after that revision.

Referee Report

3 major / 6 minor

Summary. The paper proposes Ridgeformer, a two-stage transformer architecture for cross-domain fingerprint recognition: Stage 1 extracts global features with a Vision Transformer, and Stage 2 performs fine-grained cross-sample alignment via a cross-attention module. Training combines a multi-similarity loss over three similarity matrices (CL2CL, CL2CB, CB2CB). The method is evaluated for 1:1 verification and 1:N identification on HKPolyU and RidgeBase, with an ablation study. The authors report state-of-the-art EER on HKPolyU CL2CB (2.83%) and competitive or better results on RidgeBase, and they release code.

Significance. If the claims hold, Ridgeformer would be a meaningful advance in contactless-to-contact fingerprint matching, a domain where domain shift between sensors is a known challenge. The paper has concrete strengths: it uses public benchmarks (HKPolyU, RidgeBase), reports comparisons including a COTS matcher (Verifinger), provides an ablation of the two stages, and makes code available. These features make the work reproducible in principle and of interest to the biometrics community. However, the headline claim of 'outperforms all previously benchmarked models' is not fully supported by the reported metrics, and there are open questions about the training/evaluation protocol. The central empirical contribution therefore needs revision rather than being acceptable as is.

major comments (3)

[Table 1 and Section 4.4] The claim that 'Ridgeformer outperforms all previously benchmarked models on the HKPolyU dataset' is internally contradicted by Table 1. On HKPolyU CL2CB, Ridgeformer achieves EER 2.83% and TAR@FAR=0.01 of 89.34%, while ML Fusion [20] has EER 4.07% but TAR@FAR=0.01 of 94.40%. The ROC curves cross: Ridgeformer is better at the equal-error operating point but worse at the fixed FAR=0.01 operating point, which is the more security-relevant metric reported in the same table. Section 4.4 cites only EER, ignoring the TAR metric. The conclusion should be qualified to 'lowest EER among compared methods' or the paper must justify why the FAR=0.01 operating point is not decisive.
[Section 4.1 and Section 4.2.2] The training/evaluation protocol for HKPolyU is ambiguous and potentially leaky. Section 4.2.2 states that evaluation uses the HKPolyU testing split, disjoint from training data, but Section 4.1 says 'During fine-tuning on the HKPolyU dataset [2]' the margin and threshold were changed. It is not specified whether this fine-tuning was performed on the training split only or on the full HKPolyU dataset. Furthermore, hyperparameters (alpha_pos=2.0, alpha_neg=40.0, margin, threshold tau, learning-rate decay epochs) are described as 'determined through experimentation' without stating whether the evaluation split was used for model selection. If the test split influenced fine-tuning or hyperparameter choice, the reported EER and TAR are optimistic. The authors must clarify the exact data flow.
[Table 1 and Table 2 (Ridgebase Task 1)] The claimed superiority on RidgeBase is not established because the comparison set is incomplete. For the CL2CB protocol, Table 1 lists only Verifinger and Ridgeformer; no learning-based baselines such as ML Fusion [20] or MANet [13] are reported. For the CL2CL protocol, only Verifinger and AdaCos(CNN) are compared. The abstract and Section 4.4 assert 'outperforms existing methods, including COTS solutions,' but without the relevant learned baselines on the same protocol, this is an unsupported generalization. The authors should add the missing baseline comparisons or restrict the claim to the methods actually evaluated.

minor comments (6)

[Title and Abstract] The title contains a typo: 'Mutli' should be 'Multi'.
[Section 1, Contributions] The phrase 'an multi-stage training strategy' in the second contribution should be 'a multi-stage training strategy'.
[Section 3.3, Equation formatting] The definitions of P and N in the multi-similarity loss are incomplete in the displayed equations: the subscript of S_ij is lost, making it hard to identify which pairwise similarity is being exponentiated. Please restate the equations in the standard form of [17].
[Table 2] The 100.00% Rank-1 and Rank-10 on RidgeBase CL2CL is remarkable and is not discussed. A sentence explaining why this is plausible (e.g., gallery size, high-quality images, near-perfect separation in the learned embedding) would help the reader assess the result.
[Section 4.2.1] The statement about ISPFDv2's evaluation protocol is confusing: the paper says the official protocol requires a 'random' 50-50 split with three rounds of validation, then says the authors do not report results due to non-reproducibility. It would be clearer to state explicitly that ISPFDv2 was used only for training and not for evaluation.
[References] Reference [22] is the same paper as reference [1] (Dosovitskiy et al., 'An image is worth 16x16 words'). Use a single citation with the correct conference venue.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: empirical fingerprint-matching paper with standard losses and disjoint train/test splits; the HKPolyU comparison issue is a factual inconsistency, not circularity.

full rationale

The paper contains no derivation chain that reduces to its inputs. Stage 1 and Stage 2 are architectural (ViT backbone plus cross-attention), trained with the multi-similarity loss [17] on a combined corpus; all numbers in Tables 1-3 are measured verification and identification rates on test splits. The only fitted quantities are hyperparameters (alpha_pos=2.0, alpha_neg=40.0, margin, threshold tau, learning rates), explicitly 'determined through experimentation'; these are not presented as predictions, so no fitted-input-called-prediction pattern exists. The two self-citations, RidgeBase [3] and ML Fusion [20], are dataset and baseline references used as benchmarks rather than as load-bearing justifications, and both are externally available. The internal contradiction in Section 4.4, where the paper claims Ridgeformer 'outperforms all previously benchmarked models on the HKPolyU dataset' while Table 1 shows ML Fusion has a higher TAR@FAR=0.01 (94.40%) than Ridgeformer (89.34%), is a correctness and comparison issue that can be settled directly from the table; it is not a circularity. No self-definitional equations, no imported uniqueness theorem, no ansatz smuggled in via citation, and no renamed known result appear in the manuscript. Honest non-finding: the paper is self-contained against external benchmarks and receives a score of 0.

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

The central empirical claim rests on standard deep learning assumptions and a set of tuned hyperparameters. No new entities are introduced.

free parameters (5)

alpha_pos = 2.0
Positive similarity scale in multi-similarity loss; set via experimentation in Section 4.1.
alpha_neg = 40.0
Negative similarity scale in multi-similarity loss; set via experimentation.
margin = 0.7 (Stage 1), 0.5 (Stage 2)
Margin for hard positive/negative mining; reverted to Stage 1 config during HKPolyU fine-tuning.
threshold (tau) = 0.5
Similarity threshold in loss and mining.
learning_rate_initial = 1e-5 (Stage 1), 5e-6 (fine-tune), 1e-5 (Stage 2)
Set by experimentation; decay factors 0.3 and 0.6.

assumptions (3)

domain assumption Multi-similarity loss with hard mining is an effective objective for fingerprint verification.
Adopted from [17] without verification on this task.
domain assumption A vision transformer can serve as a feature extractor for fingerprints.
The paper uses ViT as backbone, but does not specify pretraining or random init.
domain assumption The combined training set of HKPolyU, ISPFDv1/v2, and RidgeBase training splits generalizes to the test splits.
Test splits are stated to be subject-disjoint, but no cross-dataset analysis is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Ridgeformer: Mutli-Stage Contrastive Training For Fine-grained Cross-Domain Fingerprint Recognition." pith.science (2026). https://pith.science/paper/3OXS2IAZ

@misc{pith2026250601806,
  author       = {Pith},
  title        = {Pith review of: Ridgeformer: Mutli-Stage Contrastive Training For Fine-grained Cross-Domain Fingerprint Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3OXS2IAZ}},
  note         = {Machine review of arXiv:2506.01806}
}

read the original abstract

The increasing demand for hygienic and portable biometric systems has underscored the critical need for advancements in contactless fingerprint recognition. Despite its potential, this technology faces notable challenges, including out-of-focus image acquisition, reduced contrast between fingerprint ridges and valleys, variations in finger positioning, and perspective distortion. These factors significantly hinder the accuracy and reliability of contactless fingerprint matching. To address these issues, we propose a novel multi-stage transformer-based contactless fingerprint matching approach that first captures global spatial features and subsequently refines localized feature alignment across fingerprint samples. By employing a hierarchical feature extraction and matching pipeline, our method ensures fine-grained, cross-sample alignment while maintaining the robustness of global feature representation. We perform extensive evaluations on publicly available datasets such as HKPolyU and RidgeBase under different evaluation protocols, such as contactless-to-contact matching and contactless-to-contactless matching and demonstrate that our proposed approach outperforms existing methods, including COTS solutions.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

35 extracted references · 34 canonical work pages

[15]

C2cl: Contact to contactless fin- gerprint matching,

Steven A Grosz et al., “C2cl: Contact to contactless fin- gerprint matching,”IEEE Transactions on Information Forensics and Security, 2021

work page 2021
[16]

On matching finger-selfies using deep scattering networks,

Aakarsh Malhotra et al., “On matching finger-selfies using deep scattering networks,”IEEE Transactions on Biometrics, Behavior, and Identity Science, vol. 2, no. 4, 2020

work page 2020
[20]

Multi loss fusion for matching smartphone captured contactless finger images,

Bhavin Jawade et al., “Multi loss fusion for matching smartphone captured contactless finger images,” in2021 IEEE International Workshop on Information Forensics and Security (WIFS), 2021

work page 2021
[2]

To the best of our knowledge, this is the first work on contactless-to-contact fingerprint matching that em- ploys a vision transformer based architecture

work page
[13]

Matching contactless and contact- based conventional fingerprint images for biometrics identification,

Chenhao Lin et al., “Matching contactless and contact- based conventional fingerprint images for biometrics identification,”IEEE Transactions on Image Process- ing, vol. 27, no. 4, 2018

work page 2018
[1]

INTRODUCTION The demand for hygienic, portable, and robust biometric systems continues to grow, particularly in applications re- quiring secure, touchless authentication. While traditional contact-based fingerprint recognition remains dominant, it faces several challenges, including latent fingerprint theft, hygiene concerns, and the complexity of deploym...

work page
[3]

We propose an multi-stage training strategy that utilizes intra-sample cross-attention to compute fine-grained alignment score between fingerprints

work page
[4]

Ridgeformer: Mutli-Stage Contrastive Training For Fine-grained Cross-Domain Fingerprint Recognition

Experimental evaluations on the HKPolyU [2] and RidgeBase [3] datasets, demonstrating superior perfor- © 2025 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new colle...

work page Pith review arXiv 2025

Show all 35 references

[5]

Contactless MatchingWith a recent shift towards contact- less fingerprint recognition, numerous datasets have been developed [2–5]

RELA TED WORKS In this section, we will review relevant contributions in mul- tiple areas that have laid the groundwork for our proposed method. Contactless MatchingWith a recent shift towards contact- less fingerprint recognition, numerous datasets have been developed [2–5]. ...
[6]

METHOD 3.1. Problem Statement LetDbe a dataset consisting of contactless and contact- based fingerprint images from multiple subjects, represented as{V s 0 , Vs 1 , ..., Vs n }and{Q s 0, Qs 1, ..., Qs n}, whereV s i denotes thei th contactless fingerprint image of subjects, an...
[7]

Implementation Details In Stage 1 of our experiments, we trained our model using a combined training dataset

EXPERIMENTS 4.1. Implementation Details In Stage 1 of our experiments, we trained our model using a combined training dataset. For the proposed loss function, we set the positive scale(α pos)to 2.0 and the negative scale (αneg)to 40.0. The margin for mining hard positives and ...
[8]

3d fingerprint modelling and syn- thesis,

Shuqin Long et al., “3d fingerprint modelling and syn- thesis,”Electronics Letters, vol. 51, no. 18, 2015

2015
[9]

Synthetic fingerprint genera- tion,

Davide Maltoni et al., “Synthetic fingerprint genera- tion,”Handbook of fingerprint recognition, 2009

2009
[10]

Overall, fine-tuning on the HKPolyU dataset [2] provided a substantial enhance- ment of about 6% in TAR(%)@FAR= 10 −2 and a 1% de- crease in EER

resulted in a further 0.5% reduction in EER and a notable 3% increase in TAR(%)@FAR= 10−2. Overall, fine-tuning on the HKPolyU dataset [2] provided a substantial enhance- ment of about 6% in TAR(%)@FAR= 10 −2 and a 1% de- crease in EER
[11]

CONCLUSION This paper introduces a novel framework for both contactless- to-contact and contactless-to-contactless fingerprint match- ing. The framework harnesses the feature extraction capa- bilities of vision transformers to derive learned embeddings from fingerprint images ...
[14]

Ridgebase: A cross-sensor multi- finger contactless fingerprint dataset,

Bhavin Jawade et al., “Ridgebase: A cross-sensor multi- finger contactless fingerprint dataset,” in2022 IEEE International Joint Conference on Biometrics (IJCB). IEEE, 2022

2022
[17]

Syncolfinger: Synthetic contact- less fingerprint generator,

Jannis Priesnitz et al., “Syncolfinger: Synthetic contact- less fingerprint generator,”Pattern Recognition Letters, vol. 157, 2022

2022
[18]

Synthesis and reconstruction of fingerprints using generative adversarial networks,

Rafael Bouzaglo et al., “Synthesis and reconstruction of fingerprints using generative adversarial networks,” arXiv preprint arXiv:2201.06164, 2022

2022 arXiv
[19]

Arcface: Additive angular margin loss for deep face recognition,

Jiankang Deng et al., “Arcface: Additive angular margin loss for deep face recognition,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019

2019
[21]

Synthesis of multi-view 3d fingerprints to advance contactless fingerprint identifica- tion,

Chengdong Dong et al., “Synthesis of multi-view 3d fingerprints to advance contactless fingerprint identifica- tion,”IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 2023

2023
[22]

Towards more accurate contact- less fingerprint minutiae extraction and pose-invariant matching,

Hanzhuo Tan et al., “Towards more accurate contact- less fingerprint minutiae extraction and pose-invariant matching,”IEEE Transactions on Information Foren- sics and Security, vol. 15, 2020

2020
[23]

Contactless fingerprint recognition: a neural approach for perspective and ro- tation effects reduction,

Ruggero Donida Labati et al., “Contactless fingerprint recognition: a neural approach for perspective and ro- tation effects reduction,” in2013 IEEE symposium on computational intelligence in biometrics and identity management (CIBIM). IEEE, 2013

2013
[24]

Minutiae attention network with reciprocal distance loss for contactless to contact-based fingerprint identification,

Hanzhuo Tan et al., “Minutiae attention network with reciprocal distance loss for contactless to contact-based fingerprint identification,”IEEE Transactions on Infor- mation Forensics and Security, vol. 16, 2021

2021
[25]

Additionally, the contactless images in ISPFDv1 [23] were rotated to align with the orientation of the contact fingerprints

to achieve precise cropping of the fingers. Additionally, the contactless images in ISPFDv1 [23] were rotated to align with the orientation of the contact fingerprints. For the Ridgebase benchmark dataset [3], the contactless fingerprint images were similarly rotated to match ...
[26]

A cnn-based framework for com- parison of contactless to contact-based fingerprints,

Chenhao Lin et al., “A cnn-based framework for com- parison of contactless to contact-based fingerprints,” IEEE Transactions on Information Forensics and Secu- rity, vol. 14, no. 3, 2018

2018
[27]

Minutiae-guided fingerprint embeddings via vision transformers,

Steven A Grosz et al., “Minutiae-guided fingerprint embeddings via vision transformers,”arXiv preprint arXiv:2210.13994, 2022

2022 arXiv
[28]

Mra-gnn: Minutiae relation-aware model over graph neural network for fingerprint embed- ding,

Yapeng Su et al., “Mra-gnn: Minutiae relation-aware model over graph neural network for fingerprint embed- ding,” in2023 IEEE International Joint Conference on Biometrics (IJCB). IEEE, 2023

2023
[29]

Multi-similarity loss with general pair weighting for deep metric learning,

Xun Wang et al., “Multi-similarity loss with general pair weighting for deep metric learning,” inProceedings of the IEEE/CVF conference on computer vision and pat- tern recognition, 2019

2019
[30]

Dur- ing fine-tuning on the HKPolyU dataset [2], the margin and threshold were reverted to the original Stage 1 configuration, with the learning rate maintained at10 −5

Both the margin and threshold (τ) were set to 0.5. Dur- ing fine-tuning on the HKPolyU dataset [2], the margin and threshold were reverted to the original Stage 1 configuration, with the learning rate maintained at10 −5. For all experiments, we utilized the AdamW optimizer wit...
[31]

Adaface: Quality adaptive margin for face recognition,

Minchul Kim et al., “Adaface: Quality adaptive margin for face recognition,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022

2022
[32]

Fingerprint feature extraction by combining texture, minutiae, and frequency spectrum using multi-task cnn,

Ai Takahashi et al., “Fingerprint feature extraction by combining texture, minutiae, and frequency spectrum using multi-task cnn,” in2020 IEEE International Joint Conference on Biometrics (IJCB). IEEE, 2020

2020
[33]

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

Alexey Dosovitskiy et al., “An image is worth 16x16 words: Transformers for image recognition at scale,” arXiv preprint arXiv:2010.11929, 2020

2010 arXiv
[34]

On smartphone camera based fingerphoto authentication,

Anush Sankaran et al., “On smartphone camera based fingerphoto authentication,” in2015 IEEE 7th Interna- tional Conference on Biometrics Theory, Applications and Systems (BTAS), 2015

2015
[35]

Segment anything,

Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alexander C. Berg, Wan-Yen Lo, Piotr Doll ´ar, and Ross Girshick, “Segment anything,” 2023

2023
[36]

Learning transferable vi- sual models from natural language supervision,

Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever, “Learning transferable vi- sual models from natural language supervision,” 2021

2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [15]

C2cl: Contact to contactless fin- gerprint matching,

Steven A Grosz et al., “C2cl: Contact to contactless fin- gerprint matching,”IEEE Transactions on Information Forensics and Security, 2021

work page 2021

[2] [16]

On matching finger-selfies using deep scattering networks,

Aakarsh Malhotra et al., “On matching finger-selfies using deep scattering networks,”IEEE Transactions on Biometrics, Behavior, and Identity Science, vol. 2, no. 4, 2020

work page 2020

[3] [20]

Multi loss fusion for matching smartphone captured contactless finger images,

Bhavin Jawade et al., “Multi loss fusion for matching smartphone captured contactless finger images,” in2021 IEEE International Workshop on Information Forensics and Security (WIFS), 2021

work page 2021

[4] [2]

To the best of our knowledge, this is the first work on contactless-to-contact fingerprint matching that em- ploys a vision transformer based architecture

work page

[5] [13]

Matching contactless and contact- based conventional fingerprint images for biometrics identification,

Chenhao Lin et al., “Matching contactless and contact- based conventional fingerprint images for biometrics identification,”IEEE Transactions on Image Process- ing, vol. 27, no. 4, 2018

work page 2018

[6] [1]

INTRODUCTION The demand for hygienic, portable, and robust biometric systems continues to grow, particularly in applications re- quiring secure, touchless authentication. While traditional contact-based fingerprint recognition remains dominant, it faces several challenges, including latent fingerprint theft, hygiene concerns, and the complexity of deploym...

work page

[7] [3]

We propose an multi-stage training strategy that utilizes intra-sample cross-attention to compute fine-grained alignment score between fingerprints

work page

[8] [4]

Ridgeformer: Mutli-Stage Contrastive Training For Fine-grained Cross-Domain Fingerprint Recognition

Experimental evaluations on the HKPolyU [2] and RidgeBase [3] datasets, demonstrating superior perfor- © 2025 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new colle...

work page Pith review arXiv 2025