Feedback Generative Adversarial Network With Channel-Space Attention for Image-Based Optimal Path Search Planning

Path planning and seeking is one of the most challenging and interesting problems in the field of artificial intelligence. In this article, we propose a new image-based path planning algorithm to overcome traditional design concepts based on complex topological networks. Specifically, a novel feedback generative adversarial network learning model with channel-space attention (CSA-FGAN) is developed to search and plan an optimal path. It is worth noting that the input and output of the proposed feedback learning model are RGB images with complex spatial obstacle environment maps and RGB images with feasible paths, respectively. In particular, in order to further enhance the feature extraction ability and path prediction accuracy of image data, the feedback attention mechanism controls and improves the depth of the network by forming a closed loop of the proposed learning model. Final, three application scenario image examples are conducted to verify the superiority of the proposed path planning policy.