非表示:
キーワード:
-
要旨:
The question of how humans learn, represent and recognize objects has been one of the core questions in cognitive research. With the advent of the field of computer vision — most notably through the seminal work of David Marr — it seemed that the solution lay in a three-dimensional (3D) reconstruction of the environment (Marr 1982, see also one of the first computer vision systems built by Roberts et al. 1965). The success of this approach, however, was limited both in terms of explaining experimental results emerging from cognitive research as well as in enabling computer systems to recognize objects with a performance similar to humans.