Forward Head Posture Classification Using Deep Learning Models on Facial Recognition Data

International Journal on Advanced Science Engineering and Information Technology, Journal Year: 2024, Volume and Issue: 14(2), P. 805 - 810

Published: April 30, 2024

Forward Head Posture (FHP) refers to a condition where the head protrudes forward, significantly contributing neck pain and being associated with decreased productivity psychological distress. This study investigates nuanced classification of FHP proposes universally applicable methodology for its identification analysis using deep learning. Leveraging Korean Facial Image (K-FACE) dataset, rigorous image preprocessing Yolo-v8 model was conducted facilitate accurate measurement CranioVertebral Angle (CVA) from various perspectives. The meticulously evaluated effectiveness three advanced learning models: EfficientNet-B7, NFNet-F7, ResNet-152. Among these, EfficientNet-B7 demonstrated superior performance an accuracy 0.69 recall score compared other models. Additionally, comparisons based on camera angles within highlighted excellence, particularly at ±75° angle. importance regions in confirmed through Grad-CAM analysis, emphasizing critical role region accurately classifying images. comprehensive comparison proposed detailed underscore potential generalization classification. Furthermore, by leveraging unique dataset employing state-of-the-art techniques, this research offers novel perspective discourse surrounding FHP. Future could integrate expanded facial datasets apply transfer techniques further enhance precision classification, thereby improving diagnostic offering targeted interventions individuals experiencing

Language: Английский

IOTToe: A Smart System for In/Out-Toeing Foot Identification

2022 4th International Conference on Communications, Information System and Computer Engineering (CISCE), Journal Year: 2024, Volume and Issue: unknown, P. 1510 - 1514

Published: May 10, 2024

Language: Английский