Search

In brain–machine interface (BMI) systems, the performance of trained Convolutional Neural Networks (CNNs) is significantly influenced by the quality of the training data. Another issue is the training time of CNNs. This paper introduces a novel approach by combining transfer learning and a Genetic Algorithm (GA) to optimize the training data of CNNs. Transfer learning is implemented across different subjects, and the data chosen by GA aim to improve CNN performance. In addition, the GA-selected data shed light on the similarity in brain activity between subjects. Two datasets are used: (1) the publicly available BCI Competition IV, in which the subjects performed motor imagery (MI) tasks, and (2) the dataset created by healthy subjects of our laboratory performing motor movement (MO) tasks. The experimental results indicate that the brain data selected by the GA improve the recognition accuracy of the target CNN (TCNN) using pre-trained base CNN (BCNN). The improvement in accuracy is 11% and 4% for the BCI Competition IV and our laboratory datasets, respectively. In addition, the GA-selected training data reduce the CNN training time. The performance of the trained CNN, utilizing transfer learning, is tested for real-time control of a robot manipulator. [ABSTRACT FROM AUTHOR]

Problem statement: In this study, we presented a novel vision-based learning approach for autonomous robot navigation. Approach: In our method, we converted the captured image in a binary one, which after the partition is used as the input of the neural controller. Results: The neural control system, which maps the visual information to motor commands, is evolved online using real robots. Conclusion/Recommendations: We showed that evolved neural networks performed well in indoor human environments. Furthermore, we compared the performance of neural controllers with an algorithmic vision based control method. Key words: Robot navigation, vision, neural networks, evolutionary algorithm, INTRODUCTION The general problem of designing a machine for real time navigation and obstacle avoidance in an arbitrary environment is ongoing. This problem is often described as a problem in [...]

Robots operating in everyday life environments are often required to switch between different tasks. While learning and evolution have been effectively applied to single task performance, multiple task performance still lacks methods that have been demonstrated to be both reliable and efficient. This paper introduces a new method for multiple task performance based on multiobjective evolutionary algorithms, where each task is considered as a separate objective function. In order to verify the effectiveness, the proposed method is applied to evolve neural controllers for the Cyber Rodent (CR) robot that has to switch properly between two distinctly different tasks: 1) protecting another moving robot by following it closely and 2) collecting objects scattered in the environment. Furthermore, the tasks and neural complexity are analyzed by including the neural structure as a separate objective function. The simulation and experimental results using the CR robot show that the multiobjective-based evolutionary method can be applied effectively for generating neural networks that enable the robot to perform multiple tasks simultaneously. Index Terms--Evolutionary robotics, multiobjective evolution, multiple task performance, neural controller.

A multi-population genetic algorithm (MPGA) was employed to improve the performance of evolutionary algorithm, while application of an extended MPGA (EMPGA) was proposed to improve the convergence of evolution. The results indicate that EMPGA gave the best performance and evolved faster by distributing efficiently the genetic resources as different strategies became successful during the course of evolution.

Temporal and sequential information is essential to any agent continually interacting with its environment. In this paper, we test whether it is possible to evolve a recurrent neural network controller […]