Publicación: Exploring Pose Estimation with Computer Vision Processing to Model Psychomotor Performance in Karate Combats
Cargando...
Fecha
2022-09-01
Autores
Editor/a
Director/a
Tutor/a
Coordinador/a
Prologuista
Revisor/a
Ilustrador/a
Derechos de acceso
Atribución-NoComercial-SinDerivadas 4.0 Internacional
info:eu-repo/semantics/openAccess
info:eu-repo/semantics/openAccess
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Nacional de Educación a Distancia (España). Escuela Técnica Superior de Ingeniería Informática. Departamento de Inteligencia Artificial
Resumen
Technological advances enable the design of systems that interact more closely with hu-mans in a multitude of previously unsuspected fields. Martial arts are not outside the application of these techniques. From the point of view of the modeling of human movement in relation to the learning of complex motor skills, martial arts are of interest because they are articulated around a system of movements that are predefined, or at least, bounded, and governed by the laws of Physics. Their execution must be learned after continuous practice over time. Literature suggests that artifi-cial intelligence algorithms, such as those used for computer vision, can model the movements per-formed. Thus, they can be compared with a good execution as well as analyze their temporal evo-lution during learning. We are exploring the application of this approach to model psychomotor performance in karate combats (called kumites), which are characterized by the explosiveness of their movements. In addition, modeling psychomotor performance in a kumite requires the model-ing of the joint interaction of two participants, while most current research efforts in human move-ment computing focus on the modeling of movements performed individually. Thus, in this work, we explore how to apply a pose estimation algorithm to identify attack and defense movements performed by both karatekas in an ippon kihon kumite (a karate combat characterized by one-step conventional assault) and how to model their psychomotor performance. For this, we compare, us-ing an error threshold, the differences in the angles between the execution in the model (recorded in the dataset) and the current execution. These comparisons can decrease the error threshold along the evolution of the karatekas, thus allowing to measure the psychomotor learning progress. In ad-dition, postural identification of both karatekas during real kumites have also been made to confirm the viability of our proposal.
Descripción
Categorías UNESCO
Palabras clave
human activity recognition (HAR), computer vision, deep learning, human pose estima-tion (HPE), OpenPose, martial arts, karate
Citación
Centro
Facultades y escuelas::E.T.S. de Ingeniería Informática
Departamento
Inteligencia Artificial