Additive Studying in Hierarchical Neural Networks for Object Recognition

Robots that execute non-trivial projects in real-world surrounds are probably to find objects they have not seen before. Thus the power to learn new objects is a necessity skill for advanced mobile service robots. The model demonstrated in this paper has the power to learn new objects it is shown during run time. This improves the adaptability of the approach and thus enables the robot to adjust to new positions. The purpose is to assert whether and how well hierarchical neural networks are suited for this intention. The experiments conveyed to answer this question showed that the suggested incremental learning approach is applicable for hierarchical neural networks and furnishes satisfactory classification results.