The purpose of this post is to encourage more people to work on machine learning algorithms for real-world robots. End-to-end deep learning has become ubiquitous with most of today’s challenging artificial intelligence problems, such as image recognition, natural language processing, protein folding prediction and game playing agents. In fact, you would probably be laughed at for not using deep learning for these problems. But the same is not yet true for mobile robotics. Why? Deep learning has the potential to completely revolutionise mobile robotics. It understands vast quantities of data ..read more