Introduction

Everything is better with Deep Learning. Or is it? It seems that time will tell. But one undoubtable fact is that more and more issues can be solved with deep learning models. Deep Learning now plays an important role in many sciences, and computer vision isn't an exception. OpenCV has recently acquired the ability of loading and inferencing trained models from three popular frameworks: Caffe, Torch, and Tensorflow. This chapter tells you how to work with this functionality of OpenCV. This chapter also contains several useful practical applications of different existing models of classification, semantic segmentation, object detection, and other problems.