Details

Autor(en) / Beteiligte

• Defretin, J
• Marzat, Julien
• Piet-Lahanier, Helene

Titel

Learning Viewpoint Planning in Active Recognition on a Small Sampling Budget: A Kriging Approach

Ist Teil von

• 2010 Ninth International Conference on Machine Learning and Applications, 2010, p.169-174

Ort / Verlag

IEEE

Erscheinungsjahr

2010

Quelle

IEEE/IET Electronic Library (IEL)

Beschreibungen/Notizen

• This paper focuses on viewpoint planning for 3D active object recognition. The objective is to design a planning policy into a Q-learning framework with a limited number of samples. Most existing stochastic techniques are therefore inapplicable. We propose to use Kriging and bayesian Optimization coupled with Q-learning to obtain a computationally-efficient viewpoint-planning design, under a restrictive sampling budget. Experimental results on a representative database, including a comparison with classical approaches, show promising results for this strategy.