A Deep CNN-Based Feature Extraction and Matching of Pores for Fingerprint Recognition
Bibliographic record
Abstract
The inherent characteristics of fingerprint pores, including their immutability, permanence, and uniqueness in terms of size, shape, and position along ridges, make them suitable candidates for fingerprint recognition. In contrast to only a limited number of other landmarks in a fingerprint, such as minutia, the presence of a large number of pores even in a small fingerprint segment is a very attractive characteristic of pores for fingerprint recognition. A pore-based fingerprint recognition system has two main modules: a pore detection module and a pore feature extraction and matching module. The focus of this paper is on the latter module, in which the features of the detected pores in a query fingerprint are extracted, uniquely represented and then used for matching these pores with those in a template fingerprint. Fingerprint recognition systems that use convolutional neural networks (CNNs) in the design of this module have automatic feature extraction capabilities. However, CNNs used in these modules have inadequate capability of capturing deep-level features. Moreover, the pore matching part of these modules heavily relies only on the Euclidean distance metric, which if used alone, may not provide an accurate measure of similarity between the pores. In this paper, a novel pore feature extraction and matching module is presented in which a CNN architecture is proposed to generate highly representational and discriminative hierarchical features and a balance between the performance and complexity is achieved by using depthwise and depthwise separable convolutions. Furthermore, an accurate composite metric, encompassing the Euclidean distance, angle, and magnitudes difference between the vectors of pore representations, is introduced to measure the similarity between the pores of the query and template fingerprint images. Extensive experimentation is carried out to demonstrate the effectiveness of the proposed scheme in terms of performance and complexity, and its superiority over the existing state-of-the-art pore-based fingerprint recognition systems.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".