Evaluating Deep Learning Models for Autism Detection in Children Using Facial Images
Downloads
This study develops and evaluates a comprehensive deep-learning framework for early detection of Autism Spectrum Disorder (ASD) through facial image analysis. Five state-of-the-art convolutional neural network (CNN) architectures, VGG16, VGG19, ResNet50, InceptionV3, and MobileNet, were systematically assessed using a balanced dataset of 5,000 images (2,500 ASD, 2,500 non-ASD). Transfer learning and data augmentation enhanced model generalization. VGG19 achieved the highest overall accuracy (77.89%) and F1-score (0.7962), ResNet50 attained the best precision (82.53%), and InceptionV3 produced the highest recall (99.67%), indicating strong screening potential. The findings confirm that deep CNNs can capture subtle facial morphological cues linked to ASD, supporting their feasibility as non-invasive diagnostic tools. This work provides a benchmark for future multimodal, explainable, and clinically validated AI systems for autism detection.
Downloads
[1] Moridian, P., Ghassemi, N., Jafari, M., Salloum-Asfar, S., Sadeghi, D., Khodatars, M., Shoeibi, A., Khosravi, A., Ling, S. H., Subasi, A., Alizadehsani, R., Gorriz, J. M., Abdulla, S. A., & Acharya, U. R. (2022). Automatic autism spectrum disorder detection using artificial intelligence methods with MRI neuroimaging: A review. Frontiers in Molecular Neuroscience, 15, 999605. doi:10.3389/fnmol.2022.999605.
[2] Alkahtani, H., Aldhyani, T. H. H., & Alzahrani, M. Y. (2023). Deep Learning Algorithms to Identify Autism Spectrum Disorder in Children-Based Facial Landmarks. Applied Sciences (Switzerland), 13(8), 4855. doi:10.3390/app13084855.
[3] Guha, T., Yang, Z., Grossman, R. B., & Narayanan, S. S. (2018). A Computational Study of Expressive Facial Dynamics in Children with Autism. IEEE Transactions on Affective Computing, 9(1), 14–20. doi:10.1109/TAFFC.2016.2578316.
[4] Sun, B., Calvert, E. I., Ye, A., Mao, H., Liu, K., Wang, R. K., Wang, X. Y., Wu, Z. L., Wei, Z., & Kong, X. J. (2024). Interest paradigm for early identification of autism spectrum disorder: an analysis from electroencephalography combined with eye tracking. Frontiers in Neuroscience, 18(1502045). doi:10.3389/fnins.2024.1502045.
[5] Farhat, T., Akram, S., Rashid, M., Jaffar, A., Bhatti, S. M., & Iqbal, M. A. (2025). A deep learning-based ensemble for autism spectrum disorder diagnosis using facial images. PLOS One, 20(4 April), 0321697. doi:10.1371/journal.pone.0321697.
[6] Kasri, W., Himeur, Y., Copiaco, A., Mansoor, W., Albanna, A., & Eapen, V. (2025). Hybrid Vision Transformer-Mamba Framework for Autism Diagnosis via Eye-Tracking Analysis. In International Conference on Communication, Computing, Networking, and Control in Cyber-Physical Systems, CCNCPS 2025, 343–348. doi:10.1109/CCNCPS66785.2025.11135843.
[7] Vidivelli, S., Padmakumari, P., & Shanthi, P. (2025). Multimodal autism detection: Deep hybrid model with improved feature level fusion. Computer Methods and Programs in Biomedicine, 260, 108492. doi:10.1016/j.cmpb.2024.108492.
[8] Ganesh, K., Umapathy, S., & Thanaraj Krishnan, P. (2021). Deep learning techniques for automated detection of autism spectrum disorder based on thermal imaging. Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine, 235(10), 1113–1127. doi:10.1177/09544119211024778.
[9] Ahmadiar, A., Melinda, M., Muthiah, Z., Zainal, Z., & Mina Rizky, M. (2025). Thermal Image Classification of Autistic Children Using Res-Net Architecture. Indonesian Journal of Electronics, Electromedical Engineering, and Medical Informatics, 7(1), 1–10. doi:10.35882/365fkd59.
[10] Tripi, G., Roux, S., Matranga, D., Maniscalco, L., Glorioso, P., Bonnet-Brilhault, F., & Roccella, M. (2019). Cranio-facial characteristics in children with autism spectrum disorders (ASD). Journal of Clinical Medicine, 8(5), 641. doi:10.3390/jcm8050641.
[11] Sahu, R., Pattnaik, P. K., Anbananthen, K. S. M., & Muthaiyah, S. (2025). Identification of Depression Patients Using LIF Spiking Neural Network Model From the Pattern of EEG Signals. IEEE Access, 13, 55156–55168. doi:10.1109/ACCESS.2025.3552619.
[12] Monani, U. J., Samanta, S., Gourisaria, M. K., & Das, S. (2024). Efficiency Analysis of CNN through Different Filters for Medical Image Classification. 2nd IEEE International Conference on Data Science and Information System, ICDSIS 2024, 1–7. doi:10.1109/ICDSIS61070.2024.10594018.
[13] Mahmood, M. A., Jamel, L., Alturki, N., & Tawfeek, M. A. (2025). Leveraging artificial intelligence for diagnosis of children autism through facial expressions. Scientific Reports, 15(1), 8743. doi:10.1038/s41598-025-96014-6.
[14] McCarty, P., & Frye, R. E. (2020). Early Detection and Diagnosis of Autism Spectrum Disorder: Why Is It So Difficult? Seminars in Pediatric Neurology, 35, 100831. doi:10.1016/j.spen.2020.100831.
[15] Daniels, A. M., Halladay, A. K., Shih, A., Elder, L. M., & Dawson, G. (2014). Approaches to enhancing the early detection of autism spectrum disorders: A systematic review of the literature. Journal of the American Academy of Child and Adolescent Psychiatry, 53(2), 141–152. doi:10.1016/j.jaac.2013.11.002.
[16] Dawson, G., & Bernier, R. (2013). A quarter century of progress on the early detection and treatment of autism spectrum disorder. Development and Psychopathology, 25(4 Part 2), 1455–1472. doi:10.1017/S0954579413000710.
[17] Khodatars, M., Shoeibi, A., Sadeghi, D., Ghaasemi, N., Jafari, M., Moridian, P., Khadem, A., Alizadehsani, R., Zare, A., Kong, Y., Khosravi, A., Nahavandi, S., Hussain, S., Acharya, U. R., & Berk, M. (2021). Deep learning for neuroimaging-based diagnosis and rehabilitation of Autism Spectrum Disorder: A review. Computers in Biology and Medicine, 139, 104949. doi:10.1016/j.compbiomed.2021.104949.
[18] Elshoky, B. R. G., Younis, E. M. G., Ali, A. A., & Ibrahim, O. A. S. (2022). Comparing automated and non-automated machine learning for autism spectrum disorders classification using facial images. ETRI Journal, 44(4), 613–623. doi:10.4218/etrij.2021-0097.
[19] Alam, M. S., Rashid, M. M., Faizabadi, A. R., Mohd Zaki, H. F., Alam, T. E., Ali, M. S., ... & Ahsan, M. M. (2023). Efficient deep learning-based data-centric approach for autism spectrum disorder diagnosis from facial images using explainable AI. Technologies, 11(5), 115. doi:10.3390/technologies11050115.
[20] Anbananthen, S. K., Sainarayanan, G., Chekima, A., & Teo, J. (2006). Data Mining using Pruned Artificial Neural Network Tree (ANNT). Proceedings of the 2nd IEEE International Conference on Information & Communication Technologies (ICT), 1350–1356. doi:10.1109/ictta.2006.1684577.
[21] Atlam, E. S., Aljuhani, K. O., Gad, I., Abdelrahim, E. M., Atwa, A. E. M., & Ahmed, A. (2025). Automated identification of autism spectrum disorder from facial images using explainable deep learning models. Scientific Reports, 15(1), 26682. doi:10.1038/s41598-025-11847-5.
[22] K` Simon, S., Sonai Muthu Anbananthen, K., & Lee, S. (2013). A Ubiquitous Personal Health Record (uPHR) Framework. Proceedings of the 2013 International Conference on Advanced Computer Science and Electronics Information (ICACSEI 2013), 105. doi:10.2991/icacsei.2013.105.
[23] Lozier, L. M., Vanmeter, J. W., & Marsh, A. A. (2014). Impairments in facial affect recognition associated with autism spectrum disorders: A meta-analysis. Development and Psychopathology, 26(4), 933–945. doi:10.1017/S0954579414000479.
- The authors retain all copyrights. It is noticeable that authors will not be forced to sign any copyright transfer agreements.
- This work (including HTML and PDF Files) is licensed under a Creative Commons Attribution 4.0 International License.
