Deep learning practices have indicated great execution over a wide scope of fields, including natural language processing, computer vision, speech and audio processing, and robotics. These designs depend on deep neural networks, the parameterized models which utilize different layers of depiction to change information into an explicit interpretation. By utilizing unsupervised learning or self-learning paradigms, deep learning techniques can modify these systems with helpful highlights, staying away from the over-fitting issues ordinarily observed when neural networks are trained. This book enlightens the valued principles of deep learning to settle as an incredible choice for robotic applications.
Note: The entire book is written and organized in the perspective of robotics, not in the perspective of deep learning.