Efficient Discriminant Viewpoint Selection for Active Bayesian Recognition
作者:Catherine Laporte, Tal Arbel
摘要
This paper presents a novel viewpoint selection criterion for active object recognition and pose estimation whose key advantage resides in its low computational cost with respect to current popular approaches in the literature. The proposed observation selection criterion associates high utility with observations that predictably facilitate distinction between pairs of competing hypotheses by a Bayesian classifier. Rigorous experimentation of the proposed approach was conducted on two case studies, involving synthetic and real data, respectively. The results show the proposed algorithm to perform better than a random navigation strategy in terms of the amount of data required for recognition while being much faster than a strategy based on mutual information, without compromising accuracy.
论文关键词:active vision, object recognition, pose estimation, Bayesian inference, efficient viewpoint selection
论文评审过程:
论文官网地址:https://doi.org/10.1007/s11263-005-4436-9