Robot self-localization using simulated experience
Nagasaka Tomomi, Tanaka Kanji, Tetsuya Ishimaru, Ikuyo Uesaka · Society of Instrument and Control Engineers of Japan · 2010
This paper proposes a novel robust approach for robot self-localization. In our approach, realistic visual experiences are predicted in a simulated environment using a single-view 3D reconstruction as well as projection techniques. Experiments demonstrate effectiveness of the proposed approach.