Book Image

Modern Computer Vision with PyTorch

By : V Kishore Ayyadevara, Yeshwanth Reddy
Book Image

Modern Computer Vision with PyTorch

By: V Kishore Ayyadevara, Yeshwanth Reddy

Overview of this book

Deep learning is the driving force behind many recent advances in various computer vision (CV) applications. This book takes a hands-on approach to help you to solve over 50 CV problems using PyTorch1.x on real-world datasets. You’ll start by building a neural network (NN) from scratch using NumPy and PyTorch and discover best practices for tweaking its hyperparameters. You’ll then perform image classification using convolutional neural networks and transfer learning and understand how they work. As you progress, you’ll implement multiple use cases of 2D and 3D multi-object detection, segmentation, human-pose-estimation by learning about the R-CNN family, SSD, YOLO, U-Net architectures, and the Detectron2 platform. The book will also guide you in performing facial expression swapping, generating new faces, and manipulating facial expressions as you explore autoencoders and modern generative adversarial networks. You’ll learn how to combine CV with NLP techniques, such as LSTM and transformer, and RL techniques, such as Deep Q-learning, to implement OCR, image captioning, object detection, and a self-driving car agent. Finally, you'll move your NN model to production on the AWS Cloud. By the end of this book, you’ll be able to leverage modern NN architectures to solve over 50 real-world CV problems confidently.
Table of Contents (25 chapters)
Section 1 - Fundamentals of Deep Learning for Computer Vision
Section 2 - Object Classification and Detection
Section 3 - Image Manipulation
Section 4 - Combining Computer Vision with Other Techniques

Implementing a CNN

A CNN is one of the foundational blocks of computer vision techniques, and it is important for you to have a solid understanding of how they work. While we already know that a CNN constitutes convolution, pooling, flattening, and then the final classification layer, in this section, we will understand the various operations that occur during the forward pass of a CNN through code.

To gain a solid understanding of this, first, we will build a CNN architecture on a toy example using PyTorch and then match the output by building the feed-forward propagation from scratch in Python.

Building a CNN-based architecture using PyTorch

The CNN architecture will differ from the neural network architecture that we built in the previous chapter in that a CNN constitutes the following in addition to what a typical vanilla deep neural network would have:

  • Convolution operation
  • Pooling operation
  • Flattening layer

In the following code, we will build a CNN model on a toy dataset, as...