Autism Spectrum Disorder Detection Using Yolov9
Keywords:
ASD, YOLOv9, CNN, ML & DLAbstract
Autism spectrum disorder (ASD) is a neurodevelopmental illness that is marked by difficulties in social interaction, communication, and repetitive activities. It is a complex condition. For the purpose of image-based detection of autism spectrum disorder (ASD) through facial analysis, traditional deep learning techniques, particularly Convolutional Neural Networks (CNNs), have experienced widespread application. The performance of CNN-based classifiers is satisfactory; however, they frequently have a high computational cost, delayed inference, and restricted accuracy when it comes to identifying subtle facial traits that are related to autism spectrum disorder (ASD). This study presents an improved detection strategy that makes use of YOLOv9, a cutting-edge real-time object recognition model that is already well-known for its better speed and accuracy. The goal of this work is to overcome these limitations. The YOLOv9 algorithm is able to effectively recognize patterns in facial expressions that are symptomatic of autism with a much-reduced latency, which makes it acceptable for applications that include early screening. The performance of the proposed model is superior to that of standard CNN approaches because it makes use of the sophisticated feature extraction capabilities and attention mechanisms implemented in YOLOv9. The results of the experiments show that the detection precision, real-time performance, and robustness against variations in image quality and lighting have all been significantly enhanced. The findings of this study represent a significant step toward the development of autism detection techniques that are scalable, quick, and reliable in clinical and non-clinical settings.
Downloads
References
R. Haweel et al., “A Novel Grading System for Autism Severity Level Using Task-Based Functional MRI: A Response to Speech Study,” IEEE Access, vol. 9, pp. 100570–100582, 2021, doi: 10.1109/ACCESS.2021.3097606.
[2] F. Z. Subah, K. Deb, P. K. Dhar, and T. Koshiba, “A deep learning approach to predict autism spectrum disorder using multisite resting-state fmri,” Applied Sciences (Switzerland), vol. 11, no. 8, 2021, doi: 10.3390/app11083636.
[3] S. Masood, A. Srivastava, H. C. Thuwal, and M. Ahmad, “Real-time sign language gesture (word) recognition from video sequences using CNN and RNN,” Advances in Intelligent Systems and Computing, vol. 695, pp. 623–632, 2018, doi: 10.1007/978-981-107566-7_63.
[4] T. Eslami, F. Almuqhim, J. S. Raiker, and F. Saeed, “Machine learning methods for diagnosing autism spectrum disorder and attention- deficit/hyperactivity disorder using functional and structural MRI: A survey,” Frontiers in Neuroinformatics, vol. 14, 2021, doi: 10.3389/fninf.2020.575999.
[5] H. Choi, “Functional connectivity patterns of autism spectrum disorder identified by deep feature learning,” 2017, [Online]. Available: http://arxiv.org/abs/1707.07932.
[6] N. M. Rad, S. M. Kia, C. Zarbo, G. Jurman, P. Venuti, and C. Furlanello, “Stereotypical motor movement detection in dynamic feature space,” IEEE International Conference on Data Mining Workshops, ICDMW, vol. 0, pp. 487–494, 2016, doi: 10.1109/ICDMW.2016.0076.
[7] M. Havaei et al., “Brain tumor segmentation with Deep Neural Networks,” Medical Image Analysis, vol. 35, pp. 18–31, 2017, doi: 10.1016/j.media.2016.05.004.
[8] K. Devika, D. Mahapatra, R. Subramanian, and V. R. M. Oruganti, “Outlier-based autism detection using longitudinal structural MRI,” IEEE Access, vol. 10, pp. 27794–27808, 2022, doi: 10.1109/ACCESS.2022.3157613.
[9] Z. K. Khadem-Reza and H. Zare, “Automatic detection of autism spectrum disorder (ASD) in children using structural magnetic resonance imaging with machine vision system,” Middle East Current Psychiatry, vol. 29, no. 1, 2022, doi: 10.1186/s43045-02200220-1.
