TY - GEN
T1 - AutoIncSFA and vision-based developmental learning for humanoid robots
AU - Kompella, Varun Raj
AU - Pape, Leo
AU - Masci, Jonathan
AU - Frank, Mikhail
AU - Schmidhuber, Jürgen
N1 - Generated from Scopus record by KAUST IRTS on 2022-09-14
PY - 2011/12/1
Y1 - 2011/12/1
N2 - Humanoids have to deal with novel, unsupervised high-dimensional visual input streams. Our new method AutoIncSFA learns to compactly represent such complex sensory input sequences by very few meaningful features corresponding to high-level spatio-temporal abstractions, such as: a person is approaching me, or: an object was toppled. We explain the advantages of AutoIncSFA over previous related methods, and show that the compact codes greatly facilitate the task of a reinforcement learner driving the humanoid to actively explore its world like a playing baby, maximizing intrinsic curiosity reward signals for reaching states corresponding to previously unpredicted AutoIncSFA features. © 2011 IEEE.
AB - Humanoids have to deal with novel, unsupervised high-dimensional visual input streams. Our new method AutoIncSFA learns to compactly represent such complex sensory input sequences by very few meaningful features corresponding to high-level spatio-temporal abstractions, such as: a person is approaching me, or: an object was toppled. We explain the advantages of AutoIncSFA over previous related methods, and show that the compact codes greatly facilitate the task of a reinforcement learner driving the humanoid to actively explore its world like a playing baby, maximizing intrinsic curiosity reward signals for reaching states corresponding to previously unpredicted AutoIncSFA features. © 2011 IEEE.
UR - http://ieeexplore.ieee.org/document/6100865/
UR - http://www.scopus.com/inward/record.url?scp=84856318058&partnerID=8YFLogxK
U2 - 10.1109/Humanoids.2011.6100865
DO - 10.1109/Humanoids.2011.6100865
M3 - Conference contribution
SN - 9781612848679
SP - 622
EP - 629
BT - IEEE-RAS International Conference on Humanoid Robots
ER -