Learning of Object Concept Using Affinity Propagation and Logistic Regression and its Evaluation
Shohei Akimoto, Masahito Fukuda, Tomokazu Takahashi, Masato Suzuki, Yasuhiko Arai, Seiji Aoyagi · The Proceedings of Mechanical Engineering Congress Japan · 2018
In general, CNN (Convolutional Neural Networks) is used as the method with high recognition accuracy. In CNN, however, several tens of thousand images are required as learning data for each category. Also, huge learning time is required and the reason that misrecognized is not explained. In contrast, after a human just look at several objects in a category he can get something like its general object concept. Furthermore, a human can represent the concept by words. Also, a human can explain the reason that he chooses an object as an identification object. In this article, a new concept learning method based on clustering and logistic regression using object concept based on color, shape and size is proposed, which requires low dimensional multi features, small training data and short learning time. Generated object concept was evaluated in comparison with the result of recognition using real world objects included in RGB-D Object Dataset.