Unveiling The Intricacies Of Craniofacial Measurements: A Comparative Analysis Of Direct And AI/ML-Based Techniques In Cyber & Digital Forensics
Main Article Content
Abstract
Background: In the realm of digital forensics and biometrics, accurate facial measurements play a crucial role in various applications, including identification, anthropological studies, and facial reconstruction. The advent of artificial intelligence (AI) and machine learning (ML) techniques has opened up new avenues for extracting craniofacial measurements from digital photographs, offering a non-invasive and efficient alternative. However, the reliability and consistency of these AI/ML-based measurements compared to direct measurements remain an area of active research.
Aim: The primary goal of this study is to present a thorough analysis and comparison of anthropometric data gathered through caliper measurements on live subjects versus measurements derived from photographs of their frontal faces using AI and ML methods.
Methods: The research encompasses 250 diverse sample from the Indian population, aged 14 years and above, ensuring a robust and representative dataset. Fourteen craniofacial landmarks were meticulously identified for measurement and analysis.
Result: By employing advanced statistical methods such as Pearson correlation coefficients t-tests ANOVA linear regression chi-square tests as well as random forest regression techniques this study unraveled intricate patterns and correlations between the two measurement approaches.
Conclusion: In conclusion this research emphasizes the superior predictive performance of non-linear models like random forest regression in estimating live measurements based on photo-derived data indicating promising applications for AI/ML techniques in this field. Furthermore familial factors were identified to significantly influence craniofacial measurements underscoring the necessity for comprehensive modeling strategies that consider these aspects.
Article Details
References
Stephan CN, Samture S. Craniofacial identification: Techniques and forensic applications. Forensic Sci Int. 2020;308:110128. https://doi.org/10.1016/j.forsciint.2020.110128
Cesar J, Biazevic AMG, Gomes AA. Accuracy of direct and indirect anthropometric measurements for facial analysis: A systematic review. J Forensic Leg Med. 2019;62:43-9. https://doi.org/10.1016/j.jflm.2018.12.003
Nguyen HH, Bui TD. A review of deep learning for craniofacial measurement and reconstruction. IEEE Access. 2020;8:131873-97. https://doi.org/10.1109/ACCESS.2020.3010672
Stephan CN, Norris RM, Henneberg M. Does modern human cranial morphogrammetric evidence invalidate craniofacial reconstructions? Forensic Sci Int. 2005;154(2-3):195-200. https://doi.org/10.1016/j.forsciint.2004.10.004
Fourie Z, Damstra J, Gerrits PO, Ren Y. Accuracy and reliability of facial soft tissue depth measurements using cone beam computer tomography. Forensic Sci Int. 2011;211(1-3):28-34. https://doi.org/10.1016/j.forsciint.2011.04.005
Tapia JS, Álvarez C, Bermejo E, Gónzalez J, Rodríguez JM, Codorniu D. Facial automatic landmarks to estimate the measurements of a skull for forensic purposes. Forensic Sci Int. 2019;301:198-206. https://doi.org/10.1016/j.forsciint.2019.05.037
Boutellaa E, Zouba N, Bengait M, Hassiba R. Deep learning-based approach for facial landmarks detection and anthropometric measurements estimation. Sig Image Video Process. 2021;15(3):525-32. https://doi.org/10.1007/s11760-020-01821-3
Dong M, Wang L, Li Y, Zhang K, Lian J. Nasal anthropometric measurement using three-dimensional photogrammetry. J Craniofac Surg. 2019;30(7):2093-8. https://doi.org/10.1097/SCS.0000000000005732
Guyomarc'h P, Dutailly B, Charton J, Santos F. Anthropometric analysis of facial measurements and classification ethnicity in a multi-racial population. Forensic Sci Int. 2014;240:156-62. https://doi.org/10.1016/j.forsciint.2014.04.018
Jiang F, Xu C, Liu Y, Wang H, Denko M. Automated craniofacial reconstruction from skull data using deep learning. IEEE Access. 2019;7:122242-53. https://doi.org/10.1109/ACCESS.2019.2937947
Zogheib B, Majeed A, Moskowitz I, Hamarneh G. Deep learning for craniofacial landmark detection and measurement from 3D imaging data: A review. Med Image Anal. 2022;75:102273. https://doi.org/10.1016/j.media.2021.102273
Damas S, Cordón O, Ibáñez O. Forensic identification by craniofacial anthropometric analysis based on evolutionary prototype generation. Soft Comput. 2019;23(9):2975-86. https://doi.org/10.1007/s00500-018-3110-4
Guo J, Zhou J, Chen S, Le T. Craniofacial landmark detection and measurement using 3D surface data. IEEE Trans Biom Behav Identity Sci. 2021;3(1):81-93. https://doi.org/10.1109/TBIOM.2021.3058497
Shui W, Zhou M, Deng Q, Wu Z, Ji Y, Li R, et al. One millisecond face alignment with an efficient totally-anchor-free regression network. Proc IEEE/CVF Conf Comput Vis Pattern Recognit. 2020;15605-14. https://doi.org/10.1109/CVPR42600.2020.01562
Ren S, He K, Girshick R, Sun J. Faster R-CNN: Towards real-time object detection with region proposal networks. IEEE Trans Pattern Anal Mach Intell. 2017;39(6):1137-49. https://doi.org/10.1109/TPAMI.2016.2577031