A deep learning approach for forensic handwriting analysis
Abstract
Bayesian inference, which stems from Bayes’ theorem, has been the major means to identify and recognize forensic biometric (BMT) traits over the years. Parameter consideration for this theorem differs from one examiner to another based on their level of expertise and subjectivity. Issues have been raised concerning this way of identifying and recognizing BMT traits in the forensic environment; therefore, there is a need to apply deep learning models to the recognition and identification of these BMT traits. Hence, in this research, various deep learning algorithms were adopted for the classification of handwriting. The handwriting was divided into different classes. The convolutional neural network (CNN) employed for this research was trained from scratch and also off-the-shelf”, Support Vector Machine (SVM), Deep Neural Network (DNN), and Extreme Gradient Boosting (XGBoost) algorithms were also employed. Each of these algorithms performed well in various classes of these handwritings and gave varying performances in predicting the classes handwritings, with CNN having a 0.82 F-measure score and 96% accuracy leading, SVM having 79% accuracy, XGBoost having 73% accuracy, and DNN having 77% accuracy. However, CNN recorded the best result among the employed algorithms. Implicatively, CNN accurately predicted the class’s handwriting. The results obtained from this study will further assist in figuring out the factors that explain examiners’ determinations of sufficiency for individualization.
Keywords
Full Text:
PDFReferences
1. Lowe DG. Distinctive Image Features from Scale-Invariant Keypoints. International Journal of Computer Vision. 2004; 60(2): 91-110. doi: 10.1023/b: visi.0000029664.99615.94.
2. Adeyemo AB, Abiodun AO. Adaptive SIFT/SURF algorithm for off-line signature recognition. Journal of Egyptian Computer Science. 2015; 39(1): 50-56.
3. Dalal N, Triggs B. Histograms of oriented gradients for human detection. In: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2005). pp. 886-893.
4. Peterson J, Sommers I, Baskin D, Johnson D. The Role and Impact of Forensic Evidence in the Criminal Justice Process Document No.: 231977. Available online: https://www.ojp.gov/pdffiles1/nij/grants/231977.pdf (accessed on 10 October 2023).
5. Martire KA, Kemp RI, Watkins I, et al. The expression and interpretation of uncertain forensic science evidence: Verbal equivalence, evidence strength, and the weak evidence effect. Law and Human Behavior. 2013; 37(3): 197-207. doi: 10.1037/lhb0000027.
6. Thompson WC, Newman EJ. Lay understanding of forensic statistics: Evaluation of random match probabilities, likelihood ratios, and verbal equivalents. Law and Human Behavior. 2015, 39(4): 332-349. doi: 10.1037/lhb0000134.
7. Abiodun AO, Adeyemo AB. The Reproducibility and Repeatability of Modified Likelihood Ratio for Forensics Handwriting Examination. International Journal of Computer and Information Engineering. 2021; 15(5): 322-328.
8. Abiodun AO, Adeyemo AB. A Neural Network Approach to Writer’s Model for Full Likelihood Ratio in Handwriting Analysis. American Journal of Neural Networks and Applications. 2020; 6(2): 29. doi: 10.11648/j.ajnna.20200602.13
9. Abiodun AO, Adeyemo AB. An Exhaustive Mapping Model for Modified Likelihood Ratio for Handwriting Recognition in Forensic Science. Journal of Forensic Science & Criminology. 2019; 7(3): 301.
10. Chen XH, Champod C, Yang X, et al. Assessment of signature handwriting evidence via score-based likelihood ratio based on comparative measurement of relevant dynamic features. Forensic Science International. 2018; 282: 101-110. doi: 10.1016/j.forsciint.2017.11.022.
11. Abiodun OI, Jantan A, Omolara AE, et al. State-of-the-art in artificial neural network applications: A survey. Heliyon. 2018; 4(11): e00938. doi: 10.1016/j.heliyon.2018.e00938.
12. Adamu LH, Taura MG. Application of likelihood ratio and posterior probability density in sex estimation from level two fingerprint features among Hausa ethnic group. Egyptian Journal of Forensic Sciences. 2017; 7(1). doi: 10.1186/s41935-017-0026-6.
13. de Ronde A, Kokshoorn B, de Poot CJ, et al. The evaluation of finger-marks given activity level propositions. Forensic Science International. 2019; 302: 109904. doi: 10.1016/j.forsciint.2019.109904.
14. Causevic S. Frequentist vs. Bayesian Approaches in Machine Learning: A Comparison of Linear Regression and Bayesian Linear Regression. Available online: https://towardsdatascience.com/frequentist-vs-bayesian-approaches-in-machine-learning-86ece21e820e (accessed on 23 October 2024).
15. Pan Y, Huang W, Lin Z, et al. Brain tumor grading based on Neural Networks and Convolutional Neural Networks. 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). Published online August 2015. doi: 10.1109/embc.2015.7318458.
16. Shen W, Zhou M, Yang F, et al. Multi-scale Convolutional Neural Networks for Lung Nodule Classification. Information Processing in Medical Imaging. Published online 2015: 588-599. doi: 10.1007/978-3-319-19992-4_46.
17. Wolterink JM, Leiner T, Viergever MA, et al. Automatic Coronary Calcium Scoring in Cardiac CT Angiography Using Convolutional Neural Networks. Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2015. Published online 2015: 589-596. doi: 10.1007/978-3-319-24553-9_72.
18. Tistarelli M, Grosso E, Meuwly D. Biometrics in Forensic Science: Challenges, Lessons and New Technologies. Lecture Notes in Computer Science. Published online 2014: 153-164. doi: 10.1007/978-3-319-13386-7_12.
19. Gialamas DM. Criminalistics. Encyclopedia of Forensic Sciences. Published online 2000: 471-477. doi: 10.1006/rwfs.2000.0464.
20. Kasper SP. Latent Print Processing Guide. Elsevier; 2016.
21. Bleay SM, Croxton RS, Puit M. Fingerprint Development Techniques. Published online March 25, 2018. doi: 10.1002/9781119187400.
22. Daluz HM. Fingerprint Analysis Laboratory Workbook, 2nd ed. Taylor and Francis Group; 2018.
23. Wang Z, Yang J, Zhu Y. Review of Ear Biometrics. Archives of Computational Methods in Engineering. 2019, 28(1): 149-180. doi: 10.1007/s11831-019-09376-2.
24. Mishra A. Forensic Graphology: Assessment of Personality. Foresic Research & Criminology International Journal. 2017; 4(1). doi: 10.15406/frcij.2017.04.00097.
25. Fisher J, Maredia A, Nixon A, et al. Identifying Personality Traits, and Especially Traits Resulting in Violent Behavior through Automatic Handwriting Analysis. Pace University; 2012.
26. Pugnaloni M, Federiconi R. Forensic Handwriting Analysis: A Research by Means of Digital Biometrical Signature.
27. Štruc V, Žganec‐Gros J, Vesnicer B, et al. Beyond parametric score normalisation in biometric verification systems. IET Biometrics. 2014; 3(2): 62-74. doi: 10.1049/iet-bmt.2013.0076.
28. Anwar AS, Ghany KKA, ElMahdy H. Human ear recognition using SIFT features. 2015 Third World Conference on Complex Systems (WCCS). Published online November 2015. doi: 10.1109/icocs.2015.7483254.
29. Pflug A. Busch C. Ear biometrics: a survey of detection, feature extraction and recognition methods. IET Biometrics. 2012; 1(2): 114-129.
30. Zhang L, Li L, Li H, et al. 3D Ear Identification Using Block-Wise Statistics-Based Features and LC-KSVD. IEEE Transactions on Multimedia. 2016, 18(8): 1531-1541. doi: 10.1109/tmm.2016.2566578.
31. Carneiro G, Nascimento J, Bradley AP. Unregistered Multiview Mammogram Analysis with Pre-trained Deep Learning Models. Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. Published online 2015: 652-660. doi: 10.1007/978-3-319-24574-4_78.
32. Ulery BT, Hicklin RA, Buscaglia J, et al. Accuracy and reliability of forensic latent fingerprint decisions. Proceedings of the National Academy of Sciences. 2011, 108(19): 7733-7738. doi: 10.1073/pnas.1018707108.
33. Ulery BT, Hicklin RA, Buscaglia J, et al. Repeatability and Reproducibility of Decisions by Latent Fingerprint Examiners. Hsiao CK, ed. PLoS ONE. 2012, 7(3): e32800. doi: 10.1371/journal.pone.0032800.
34. Ulery BT, Hicklin RA, Kiebuzinski GI, et al. Understanding the sufficiency of information for latent fingerprint value determinations. Forensic Science International. 2013, 230(1-3): 99-106. doi: 10.1016/j.forsciint.2013.01.012.
35. Ulery BT, Hicklin RA, Roberts MA, Buscaglia J. Measuring what latent fingerprint examiners consider sufficient information for individualization determinations. PLoS One. 2014; 9: e110179.
36. Haraksim, R. Meuwly, D. Fingerprint Evidence Evaluation – Robustness to the Lack of Data, in proceedings EAFS 2012, Den Haag.
37. Haraksim R, Meuwly D, Doekhie G, et al. Assignment of evidential value of a fingermark general pattern using a Bayesian Network. BIOSIG; 2013.
38. Liu S, Champod C, Wu J, Luo Y. Study on Accuracy of Judgments by Chinese Fingerprint Examiners. Journal of Forensic Science and Medicine. 2015; 1(1): 33. doi: 10.4103/2349-5014.157908.
39. Guo G, Ray A, Izydorczak M, et al. Unveiling intra-person fingerprint similarity via deep contrastive learning. Science Advances. 2024, 10(2). doi: 10.1126/sciadv.adi0329.
40. Arbab-Zavar B, Wei X, Bustard JD, et al. On Forensic Use of Biometrics. Handbook of Digital Forensics of Multimedia Data and Devices. Published online July 6, 2015: 270-304. doi: 10.1002/9781118705773.ch7.
41. Wei X, Li CT, Hu Y. Face recognition with occlusion using Dynamic Image-to-Class Warping (DICW). 2013 10th IEEE International Conference and Workshops on Automatic Face and Gesture Recognition (FG). Published online April 2013. doi: 10.1109/fg.2013.6553747.
42. Reid DA, Nixon MS, Stevenage SV. Soft Biometrics, Human Identification Using Comparative Descriptions. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2013; 36(6): 1216-1228. doi: 10.1109/tpami.2013.219.
43. Alsaadi IM. ‘‘Physiological Biometric Authentication Systems, Advantages, Disadvantages and Future Development: A Review’’ International Journal of Scientific and Technology Research. 2015, 4(12).
44. Kavitha VP, Priyatha M. A Novel Digital Image Forgery Detection Method Using Svm Classifier. International Journal of Advanced Research in Electrical, Electronics and Instrumentation Engineering. 2017; 3(2): 56-71.
45. Asanga MP, Essiet UU, Ukhurebor KE, et al. Social media and Academic Performance: A Survey Research of Senior Secondary School Students in Uyo, Nigeria. International Journal of Learning, Teaching and Educational Research. 2023; 22(2): 323-337. doi: 10.26803/ijlter.22.2.18.
46. Hussain A, Sipra HFK, Waheed A, et al. Exploring the academic perceptions of climate engineering in developing countries. ATMÓSFERA. Published online November 21, 2023. doi: 10.20937/atm.53264.
47. Hussaini AR, Ibrahim S, Ukhurebor KE, et al. The Influence of Information and Communication Technology in the Teaching and Learning of Physics. International Journal of Learning, Teaching and Educational Research. 2023; 22(6): 98-120. doi: 10.26803/ijlter.22.6.6.
48. Ndunagu JN, Ukhurebor KE, Adesina A. Virtual Laboratories for STEM in Nigerian Higher Education: The National Open University of Nigeria Learners’ Perspective. In: Elmoazen R, López-Pernas S, Misiejuk K, et al. (editors). Technology-Enhanced Learning in Laboratories Workshop; 2023.
49. Nneji CC, Urenyere R, Ukhurebor KE, et al. The impacts of COVID-19-induced online lectures on the teaching and learning process: An inquiring study of junior secondary schools in Orlu, Nigeria. Frontiers in Public Health. 2022, 10. doi: 10.3389/fpubh.2022.1054536.
50. Nwankwo W, Ukhurebor KE. Web Forum and social media: A Model for Automatic Removal of Fake Media using Multilayered Neural Networks. International Journal of Scientific & Technology Research. 2020; 9(1): 4371-4377.
51. Bansal D, Kaushal S. A Novel Analysis of Image Forgery Detection Using SVM. International Journal of Engineering and Applied Sciences. 2016; 3(12).
52. Lund SP, Iyer H. Likelihood Ratio as Weight of Forensic Evidence: A Closer Look. Journal of Research of the National Institute of Standards and Technology. 2017; 122. doi: 10.6028/jres.122.027.
53. Kavitha VP, Priyatha M. A Novel Digital Image Forgery Detection Method Using SVM Classifier. IJAREEIE. 2014; 3(2): 56-71.
54. Billah M, Waheed S. Minimum redundancy maximum relevance (mRMR) based feature selection from endoscopic images for automatic gastrointestinal polyp detection. Multimedia Tools and Applications. 2020; 79(33-34): 23633-23643. doi: 10.1007/s11042-020-09151-7.
55. Islam MT, Islam MR, Uddin MP, et al. A Deep Learning-Based Hyperspectral Object Classification Approach via Imbalanced Training Samples Handling. Remote Sensing. 2023; 15(14): 3532. doi: 10.3390/rs15143532.
56. Hayashi T, Gyohten K, Ohki H, et al. A Study of Data Augmentation for Handwritten Character Recognition using Deep Learning. 2018 16th International Conference on Frontiers in Handwriting Recognition (ICFHR). Published online August 2018. doi: 10.1109/icfhr-2018.2018.00102.
57. Zhou W, Gao S, Zhang L, et al. Histogram of Oriented Gradients Feature Extraction From Raw Bayer Pattern Images. IEEE Transactions on Circuits and Systems II: Express Briefs. 2020, 67(5): 946-950. doi: 10.1109/tcsii.2020.2980557.
58. Alfalou A, Jridi M, Kourtli Y, et al. A comparative study of CFs, LBP, HOG, SIFT, SURF, and BRIEF techniques for face recognition. Alam MS, ed. Pattern Recognition and Tracking XXIX. Published online April 30, 2018. doi: 10.1117/12.2309454.
59. Huang S, Nianguang CAI, Penzuti Pacheco P, et al. Applications of support vector machine (SVM) learning in cancer genomics. Cancer Genomics and Proteomics. 2018; 15(1): 41-51.
60. Cao J, Wang M, Li Y, et al. Improved support vector machine classification algorithm based on adaptive feature weight updating in the Hadoop cluster environment. Bagci U, ed. PLOS ONE. 2019; 14(4): e0215136. doi: 10.1371/journal.pone.0215136.
61. Chen T, Guestrin C. XGBoost. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Published online August 13, 2016. doi: 10.1145/2939672.2939785.
62. Indolia S, Goswami AK, Mishra SP, et al. Conceptual Understanding of Convolutional Neural Network—A Deep Learning Approach. Procedia Computer Science. 2018; 132: 679-688. doi: 10.1016/j.procs.2018.05.069.
63. Tan M, Le QV. EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. Available online: http://proceedings.mlr.press/v97/tan19a.html (accessed on 10 October 2023).
DOI: https://doi.org/10.32629/jai.v7i5.1533
Refbacks
- There are currently no refbacks.
Copyright (c) 2024 Adeyinka Oluwabusayo Abiodun, Olalekan M. Awoniran, Kingsley Eghonghon Ukhurebor, Ozichi Nweke Emuoyibofarhe, Adetoye Adeyemo, Idemudia Edetalehn Oaihimire
License URL: https://creativecommons.org/licenses/by-nc/4.0/