Title :
Qualitative Evaluation of Full Body Movements with Gesture Description Language
Author :
Hachaj, Tomasz ; Ogiela, Marek R.
Author_Institution :
Inst. of Comput. Sci. & Comput. Methods, Pedagogical Univ. of Krakow, Krakow, Poland
Abstract :
The motivation for writing this paper is to present the method for qualitative evaluation of full body movements with Gesture Description Language. By qualitative evaluation we understand the possibility of evaluating the similarity between the template of movement defined in GDL script (GDLs) and physical activity performed by observed user. The establishing of this methodology allows us to perform real time evaluation of movements accuracy defined as similarity to the statistically computed templates. The presented methodology can be used not only to perform advanced numerical analysis of movements but also to support multimedia visual feedback to various movements activities in computer applications. The initial evaluation presented in this paper proved that proposed method of qualitative analysis of movements has a potential to be used in practice.
Keywords :
gesture recognition; specification languages; GDL script; full body movement; gesture description language; multimedia visual feedback; numerical analysis; qualitative analysis; qualitative evaluation; Elbow; Hidden Markov models; Hip; Joints; Manganese; Multimedia communication; Real-time systems; Gesture Description Language; gestures recognition; movements recognition; qualitative evaluation; unsupervised learning;
Conference_Titel :
Artificial Intelligence, Modelling and Simulation (AIMS), 2014 2nd International Conference on
Print_ISBN :
978-1-4799-7599-0
DOI :
10.1109/AIMS.2014.32
