… of optimal viewpoint selection for 3D object recognition without making a priori assumptions
about the objects and the classifier. The problem is to determine the next view of an object …

… In section 2 we will show how the viewpoint fusion of multiple views can be done … the object
recognition part of our solution. The reinforcement learning approach for viewpoint selection …

… Object recognition problems in computer vision are often based on single image data pro…
for active object recognition, where a camera is selectively moved around a considered object. …

… Constructing an aspect graph requires partitioning the viewpoint space into view-equivalent
cells by a number of visual event surfaces [1], [2], [3], [4]. However, construction of an aspect …

… views (viewpoint selection) for 3-D object recognition. We formally define the selection of
additional views … to use reinforcement learning for continuous viewpoint training and selection …

… selection can be trained unsupervised and can handle continuous viewpoints and pose …
object recognition approach. In section 3 we give an overview of our active viewpoint selection …

… solve a particular We present an active object recognition strategy which com- vision task. …
to for a 3D object in a 2D image, with a viewpoint control strategy search for a particular object. …

… 3D object recognition based on multi-view for the first time, excluding other 3D object recognition
… Active camera viewpoints selection methods can achieve object recognition with as few …

… objects based on multiple spatially-registered images of a scene. We argue that such a
multi-view … robotics and surveillance than traditional object recognition where only a single image …

… and viewpoint selection actions. In the training step for each (pan, tilt, zoom) setting a, we took
views from each object … of 100 new disturbed views for each object class and projected the …