In this paper, we address the issue robots developing a reaching ability, as human infants do. Specifically, a novel infant-inspired framework is proposed based on recent findings that the emergence of reaching is the product of a deeply embodied process. The methodology behind our framework is to develop proprioception, fixation, and internal models (including both the forward and inverse models) in three sequential phases, with the whole developing process being driven by self-produced motor babbling. In the first phase, an autoencoder-based proprioception model is proposed. In the second phase, a simplified strategy for imitating the function of fixation is developed. In the third phase, a new forward model and two new inverse models are further proposed. We evaluated our framework and associated models with the PKU-HR6.0II physical robot and two simulated robots. Experiments confirm that the PKU-HR6.0II could successfully develop its reaching ability in a manner similar to that of human infants through our proposed framework. Better performance and adaptability are also demonstrated in comparison to existing benchmarks. It is also shown that the proposed framework has the potential to achieve other manipulation abilities such as grasping and placing.
ABSTRACT
PUBLICATION RECORD
- Publication year
2018
- Venue
IEEE Transactions on Cognitive and Developmental Systems
- Publication date
2018-09-01
- Fields of study
Computer Science, Engineering
- Identifiers
- External record
- Source metadata
Semantic Scholar
CITATION MAP
EXTRACTION MAP
CLAIMS
- No claims are published for this paper.
CONCEPTS
- No concepts are published for this paper.
REFERENCES
Showing 1-42 of 42 references · Page 1 of 1
CITED BY
Showing 1-9 of 9 citing papers · Page 1 of 1