Preliminary Investigation on Visual Finger-Counting with the iCub Robot Cameras and Hands
This short paper describes an approach for collecting a dataset of hand’s pictures and training a Deep Learning network that could enable the iCub robot to count on its fingers using solely its own cameras. Such a skill, mimicking children’s habits, can support arithmetic learning in a baby robot, an important step in creating artificial intelligence for robots that could learn like children in the context of cognitive developmental robotics. Preliminary results show the approach is promising in terms of accuracy.
KeywordsDevelopmental robotics Finger-counting Faster R-CNN iCub
This work has been supported by the EPSRC through the grant no. EP/P030033/1 (NUMBERS). Authors are grateful to the NVIDIA Corporation for donating GeForce GTX TITAN X that has been used to accelerate the computation.
- 4.Fischer, M.H., et al.: Finger Counting and Numerical Cognition. Front. Psychol. 3, 108 (2012)Google Scholar
- 5.Goldin-Meadow, S., et al.: Gesture’s role in learning arithmetic. In: Edwards, L.D., et al. (eds.) Emerging Perspectives on Gesture and Embodiment in Mathematics, pp. 51–72. Information Age Publishing, Charlotte (2014)Google Scholar
- 6.De La Cruz, V.M. et al.: Making fingers and words count in a cognitive robot. Front. Behav. Neurosci. 8, 12 pages (2014)Google Scholar
- 7.Leitner, J. et al.: Humanoid learns to detect its own hands. In: 2013 IEEE Congress Evolutionary Computation, CEC 2013, pp. 1411–1418 (2013)Google Scholar
- 8.Di Nuovo, A., Jay, T.: The development of numerical cognition in children and artificial systems: a review of the current knowledge and proposals for multi-disciplinary research. IET Cogn. Comput. Syst. 1, 2–11 (2019)Google Scholar