This paper presents a survey on datasets created for the field of gesture recognition. The main characteristics of the datasets are presented on two tables to provide researchers a clear and rapid access to the information. This paper also provides a comprehensive description of the datasets and discusses their general strengths and limitations. Guidelines for creation and selection of datasets for gesture recognition are proposed. This survey should be a key-access point for researchers looking to create or use datasets in the field of human gesture recognition.
Détails
Titre
A Survey of datasets for human gesture recognition
Auteur(s)/ trice(s)
Ruffieux, Simon (School of Engineering and Architecture (HEIA-FR), HES-SO University of Applied Sciences and Arts Western Switzerland) Lalanne, Denis (University of Fribourg, Fribourg, Switzerland) Mugellini, Elena (School of Engineering and Architecture (HEIA-FR), HES-SO University of Applied Sciences and Arts Western Switzerland) Khaled, Omar Abou (School of Engineering and Architecture (HEIA-FR), HES-SO University of Applied Sciences and Arts Western Switzerland)
Date
2014-06
Publié dans
Proceedings of the 6th International Conference HCI International 2014, 22-27 June 2014, Heraklion, Crete, Greece
Editeur
Heraklion, Crete, Greece, 22-27 June 2014
Pagination
pp. 337-348
Présenté à
Human-Computer Interaction. Advanced Interaction Modalities and Techniques (HCI 2014), Heraklion, Crete, Greece, 2014-06-22, 2014-06-27