International Journal of Computer Science & Network Security

International Journal of Computer Science & Network Security (국제컴퓨터통신보호논문지학회)
권호 표지
DOI QR코드

DOI QR Code

Exploring the Feasibility of Neural Networks for Criminal Propensity Detection through Facial Features Analysis

  • Amal Alshahrani (College of Computing, Computer Science and Artificial Intelligence Department, Umm Al Qura University) ;
  • Sumayyah Albarakati (College of Computing, Computer Science and Artificial Intelligence Department, Umm Al Qura University) ;
  • Reyouf Wasil (College of Computing, Computer Science and Artificial Intelligence Department, Umm Al Qura University) ;
  • Hanan Farouquee (College of Computing, Computer Science and Artificial Intelligence Department, Umm Al Qura University) ;
  • Maryam Alobthani (College of Computing, Computer Science and Artificial Intelligence Department, Umm Al Qura University) ;
  • Someah Al-Qarni (College of Computing, Computer Science and Artificial Intelligence Department, Umm Al Qura University)
  • Received : 2024.05.05
  • Published : 2024.05.30
Download PDF
⟨ Previous Next ⟩

Abstract

While artificial neural networks are adept at identifying patterns, they can struggle to distinguish between actual correlations and false associations between extracted facial features and criminal behavior within the training data. These associations may not indicate causal connections. Socioeconomic factors, ethnicity, or even chance occurrences in the data can influence both facial features and criminal activity. Consequently, the artificial neural network might identify linked features without understanding the underlying cause. This raises concerns about incorrect linkages and potential misclassification of individuals based on features unrelated to criminal tendencies. To address this challenge, we propose a novel region-based training approach for artificial neural networks focused on criminal propensity detection. Instead of solely relying on overall facial recognition, the network would systematically analyze each facial feature in isolation. This fine-grained approach would enable the network to identify which specific features hold the strongest correlations with criminal activity within the training data. By focusing on these key features, the network can be optimized for more accurate and reliable criminal propensity prediction. This study examines the effectiveness of various algorithms for criminal propensity classification. We evaluate YOLO versions YOLOv5 and YOLOv8 alongside VGG-16. Our findings indicate that YOLO achieved the highest accuracy 0.93 in classifying criminal and non-criminal facial features. While these results are promising, we acknowledge the need for further research on bias and misclassification in criminal justice applications

Keywords

References

  1. Wang, M., & Deng, W. (2020). Deep learning for criminal justice reform. Nature Machine Intelligence, 2(1), 11-18. https://www.nature.com/articles/nature14539
  2. Bias in Bios: Fairness, Accountability, and Transparency in Biometric Systems. (2019). National Academies Press. https://dl.acm.org/doi/abs/10.1145/3287560.3287572
  3. Garvie, C. (2018). Examining the impact of algorithmic bias Criminology, 108(4), 825-870. https://www.bu.edu/articles/2023/do-algorithms-reducebias-in-criminal-justice/
  4. O'Neil, C. (2017). Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy. Broadway Books.
  5. Valla, J. M., Ceci, S. J., & Williams, W. M. (2011). The Accuracy of Inferences About Criminality Based on Facial Appearance. Journal of Social, Evolutionary, and Cultural Psychology, 5(1), 66-91.
  6. X. Wu, X. Zhang. "Automated inference on criminality using face images," in arXiv preprint arXiv:1611.04135, pp. 4038-4052, 2016.
  7. R. Ranjan, S. Sankaranarayanan, C. D. Castillo and R. Chellappa, "An All-In-One Convolutional Neural Network for Face Analysis," 2017 12th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2017), Washington, DC, USA, 2017, pp. 17-24, doi: 10.1109/FG.2017.137.
  8. Johnson, H., Anderson, M., Westra, H. R., & Suter, H. (2018). Inferences on Criminality Based on Appearance. SciSpace - Paper. https://typeset.io/papers/inferences-on-criminality-based-on-appearance-vbtlba1570
  9. M. Hashemi and M. Hall, 'RETRACTED ARTICLE: Criminal tendency detection from facial images and the gender bias effect', Journal of Big Data, vol. 7, no. 1, p. 2, 2020.
  10. K. W. Bowyer, M. C. King, W. J. Scheirer, and K. Vangara, 'The "criminality from face" illusion', IEEE Transactions on Technology and Society, vol. 1, no. 4, pp. 175-183, 2020. https://doi.org/10.1109/TTS.2020.3032321
  11. Sheldon, K. M., Corcoran, M., & Trent, J. (2020). The face of crime: Apparent happiness differentiates criminal and non-criminal photos. The Journal of Positive Psychology, 1-18. doi:10.1080/17439760.2020.1805500
  12. Keles, U., Lin, C. & Adolphs, R. A Cautionary Note on Predicting Social Judgments from Faces with Deep Neural Networks. Affec Sci 2, 438-454 (2021). https://doi.org/10.1007/s42761-021-00075-5
  13. Rasmussen, S.H.R., Ludeke, S.G. & Klemmensen, R. Using deep learning to predict ideology from facial photographs: expressions, beauty, and extra-facial information. Sci Rep 13, 5257 (2023).
  14. G. James, P. Okafor, E. Chukwu, N. Michael, and O. Ebong, "Predictions of Criminal Tendency Through Facial Expression Using Convolutional Neural Network", journalisi, vol. 6, no. 1, pp. 13-29, Mar. 2024. https://doi.org/10.51519/journalisi.v6i1.635
  15. "Illinois DOC labeled faces dataset", www.kaggle.com. https://www.kaggle.com/datasets/davidjfisher/illinois-doc-labeled-faces-dataset?resource=download
  16. C. E. Thomaz and G. A. Giraldi. A new ranking method for Principal Components Analysis and its application to face image analysis, Image and Vision Computing, vol. 28, no. 6, pp. 902-913, June 2010. https://doi.org/10.1016/j.imavis.2009.11.005
  17. Kenhub. Regions of the Head and Neck. Retrieved April 27, 2024, from https://www.kenhub.com/en/library/anatomy/regions-of-the-head-and-neck
  18. J. Redmon, "YOLO: Real-Time Object Detection," Pjreddie.com, 2012. https://pjreddie.com/darknet/yolo/
  19. K. Team, "Keras documentation: VGG16 and VGG19," keras.io. https://keras.io/api/applications/vgg/
  20. Kingma, D. P., & Ba, J. L. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.
  21. Micikevicius, P., Jouppi, N., Kashuk, A., Anguena, J., Tensor Processing Unit Architecture, (2017). Communications of the ACM, 61(7), 10-18. https://doi.org/10.1145/3168260
  22. "VGG-16 convolutional neural network - MATLAB vgg16," www.mathworks.com.https://www.mathworks.com/help/deeplearning/ref/vgg16.html