Comparative study of machine learning algorithms for face recognition
Bibliographic record
Abstract
Background: The fundamental need for authentication and identification of humans using their physiological, behavioral or biological characteristics, continues to be applied extensively to secure localities, property, financial transactions, etc. Biometric systems based on face characteristics, continue to attract the attention of researchers, major public and private services. In the literature, many methods have been deployed by different authors. The best performance must be found in order to be able to recommend the most effective method. So, the main objective of thisarticle is to make a comparative study of different existing techniques.Methods: A biometric system is generally composed of four stages: acquisition of facial images, preprocessing, extraction of characteristics and finally classification. In this work, the focus is on machine learning algorithms for classification. These algorithms are: Support Vector Machines (SVM), Artificial Neural Networks (ANN), K-Nearest Neighbors (KNN), Random Forests (RF), Logistic Regression (LR), Naive Bayesian Classification (NB: Naive Bayes’ Classifiers) and deep learning techniques such as Convolutional Neural Networks (CNN). The comparison criterion is the average performance, calculated using three performance measures: recognition rate, confusion matrix, and the Area Under Receiver Operating Characteristic (ROC) curve.Results: Based on this criterion, the performance comparison of selected machine learning algorithms, shows that CNN is the best, with an average performance of 100.00% On ORL face database. However, on the YALE database, classical algorithms such as artificial neural networks have obtained the best performances, the highest being a rate of 100%.Discussion: Deep learning techniques are very efficient in image classification as proven by the results on the ORL database. However, their inefficiency on YALE face database is due to the small size of this database which is inappropriate for some deep learning algorithms. But this weakness can be corrected by image augmentation techniques. The comparison of these results with existing state-of-the-art methods is nearly the same. Authors achieved performances of 94.82%, 95.79%, 96.15%, 96.44%, 97.27%, 98.52% and 98.95% for NB, KNN, RF, LR, ANN, SVM and CNN classifiers, respectively. Finally, in depth discussion, it is concluded that between all these approaches which are useful in face recognition, the CNN is the best classification algorithm.
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 | 0.005 | 0.012 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".