Summary

In this chapter, you learned about Transfer Learning. We explored different libraries and approaches in order to build a real-time object detection application. You learned how to set up OpenCV and looked at how it is rather useful in building the baseline application. In this baseline approach, we used the model that is trained using the caffe deep learning library. After that, we used TensorFlow to build real-time object detection, but in the end, we used a pre-trained YOLO model, which outperformed every other approach. This YOLO-based approach gave us more generalized approach for object detection applications. If you are interested in building innovative solutions for computer vision, then you can enroll yourself in the VOC challenges. This boosts your skills and gives you a chance to learn. You can refer to this link for more information: http://host.robots.ox.ac.uk/pascal/VOC/ (PASCAL VOC Challenges 2005-2012). You can also build your own algorithm and check the result and...