3D Deep Learning with Python: Design and develop your computer vision model with 3D data using PyTorch3D and more

by Xudong Ma, Vishakh Hegde, Lilit Yolyan

Overview

With this hands-on guide to 3D deep learning, developers working with 3D computer vision will be able to put their knowledge to work and get up and running in no time.
Complete with step-by-step explanations of essential concepts and practical examples, this book lets you explore and gain a thorough understanding of state-of-the-art 3D deep learning. You’ll see how to use PyTorch3D for basic 3D mesh and point cloud data processing, including loading and saving ply and obj files, projecting 3D points into camera coordination using perspective camera models or orthographic camera models, rendering point clouds and meshes to images, and much more. As you implement some of the latest 3D deep learning algorithms, such as differential rendering, Nerf, synsin, and mesh RCNN, you’ll realize how coding for these deep learning models becomes easier using the PyTorch3D library.
By the end of this deep learning book, you’ll be ready to implement your own 3D deep learning models confidently.

Product Details

ISBN-13:	9781803233680
Publisher:	Packt Publishing
Publication date:	10/31/2022
Sold by:	Barnes & Noble
Format:	eBook
Pages:	236
File size:	9 MB

About the Author

Xudong Ma is a Staff Machine Learning engineer with Grabango Inc. at Berkeley California. He was a Senior Machine Learning Engineer at Facebook(Meta) Oculus and worked closely with the 3D PyTorch Team on 3D facial tracking projects. He has many years of experience working on computer vision, machine learning and deep learning. He holds a Ph.D. in Electrical and Computer Engineering.


Vishakh Hegde is a Machine Learning and Computer Vision researcher. He has over 7 years of experience in this field during which he has authored multiple well cited research papers and published patents. He holds a masters from Stanford University specializing in applied mathematics and machine learning, and a BS and MS in Physics from IIT Madras. He previously worked at Schlumberger and Matroid. He is a Senior Applied Scientist at Ambient.ai, where he helped build their weapon detection system which is deployed at several Global Fortune 500 companies. He is now leveraging his expertise and passion to solve business challenges to build a technology startup in Silicon Valley. You can learn more about him on his personal website.


Lilit Yolyan is a machine learning researcher working on her Ph.D. at YSU. Her research focuses on building computer vision solutions for smart cities using remote sensing data. She has 5 years of experience in the field of computer vision and has worked on a complex driver safety solution to be deployed by many well-known car manufacturing companies.

Table of Contents

Table of Contents

1. 3D data file formats - ply and obj, 3D coordination systems, camera models
2. Basic rendering concepts, basic PyTorch optimization, heterogeneous batching
3. Fitting using deformable mesh models
4. Differentiable rendering basic concepts
5. Differentiable volume rendering
6. NeRF - Neural Radiance Fields
7. GIRAFFE
8. Human body 3D fitting using SMPL models
9. Synsin - end-to-end view synthesis from a single image
10. Mesh RCNN