Due to the DANA that affected the Valencian Community and its terrible consequences, the ValgrAI team decided not to hold the XII Research Morning Meeting on its scheduled date. However, with renewed enthusiasm and commitment, we have conducted the XIII Research Morning Meeting as the last meeting of the year, reaffirming our commitment to the development and dissemination of knowledge in the field of artificial intelligence.
Featured Presentations
Evaluation of Balance Postures in the Short Physical Performance Battery Using Mobile Device Cameras
Arturo Gascó presented research focused on the digitalization of balance tests from the Short Physical Performance Battery (SPPB) using deep learning models integrated into a mobile application. Leveraging the accessibility of RGB cameras in mobile devices, the study evaluated the performance of several models, including customized CNNs, ResNet50, VGG16, MobileNetV3Small, Vision Transformers, and YOLOv8, for classifying balance postures.
The results showed that YOLOv8, especially in its lightweight versions, offers an optimal balance between accuracy and speed, making it suitable for real-time implementation on mobile platforms. Processes such as data augmentation and manual segmentation were essential for training robust models despite a limited dataset. The study concludes that this approach provides an accessible tool for frailty assessment and highlights the need for further optimization and data expansion for broader applicability in real-world environments.
Next event
The ValgrAI team thanks all participants, speakers, and attendees for their continued support throughout the year. We look forward to more innovative research and collaborations in the next cycle of Research Morning Meeting starting in January 2025. For more information and future registrations, contact us at info@valgrai.eu.
See you in 2025 to continue exploring the fascinating world of artificial intelligence!
