Summary

In this chapter, we presented a new way of looking at the object detection task. The 3D world requires solutions that work accordingly, and this is one of the first approaches toward that goal. We learned how Mesh R-CNN works by understanding the architecture and the structure of the model. We dove deeper into some interesting operations and techniques that are used in the model, such as graph convolutional networks, Cubify operations, the mesh predictor structure, and more. Finally, we learned how this model can be used in practice to detect objects on the image that the network has never seen before. We evaluated the results by rendering the 3D object.

Throughout this book, we have covered 3D deep learning concepts, from the basics to more advanced solutions. First, we learned about the various 3D data types and structures. Then, we delved into different types of models that solve different types of problems such as mesh detection, view synthesis, and more. In addition...