Skeleton-based action recognition is becoming more widely used in Human-computer Interaction (HCI) due to its unique and intuitive way of controlling applications in various fields. This research project was designed around creating a new shape retrieval challenge (SHREC) for the hand gestures category. SHREC's general objective is to create 3D Shape Retrieval Challenges to evaluate 3D shape retrieval systems in the hope of finding new ways on how we can improve these retrieval systems.
In this thesis, we address the challenges of creating a dataset of hand motions molding clay captured using a Vicon system, explain all steps involved in creating the skeletal coordinate system from the Vicon data, and present the steps we took to create a new SHREC track for the hand gesture category. We also explain the creation and evaluation of our skeleton-based recognition system.
Our contest will focus on trying to recognize highly similar hand motions from a professional potter. Participants will try to recognize the hand motions from the coordinate data of both potter's hands. This track is made to see if hand recognition systems can recognize similar hand motions using both hands. These motions are all recorded using a Vicon system and pre-processed into a coordinate system using Blender and Vicon Shogun Post. For more information about our SHREC challenge please look at our SHREC page: https://www.shrec.net/SHREC-2024-hand-motion/, and the following paper:
Ben Veldhuijzen, Remco C. Veltkamp, Omar Ikne, Benjamin Allaert, Hazem Wannous, Marco Emporio, Andrea Giachetti, Joseph J. LaViola Jr., Ruiwen He, Halim Benhabiles, Adnane Cabani, Anthony Fleury, Karim Hammoudi, Konstantinos Gavalas, Christoforos Vlachos, Athanasios Papanikolaou, Ioannis Romanelis, Vlassis Fotis, Gerasimos Arvanitis, Konstantinos Moustakas, Martin Hanik, Esfandiar Nava-Yazdani, Christoph von Tycowicz.
SHREC 2024: Recognition of dynamic hand motions molding clay.
Computers & Graphics 123, 2024 DOI 10.1016/j.cag.2024.104012