Highlights: In this post, we will see how we can create a face model just from an input image using PyTorch. For this 3D model, we will obtain all the parameters which we can later play with. So, let’s begin with our post. Data preparation The first step for running the code is cloning the GitHub repository from Ascust, which has very well-explained instructions for running the code. The next step is to download the…
Read more
Highlights: In this post, we are going to learn several hacks that we can use to train stable GAN models. First, we are going to provide a quick recap of the GANs theory, and then, we are going to talk about challenges when training GANs. After that, we will provide solutions for these challenges in Python. So, let’s begin with our post. Tutorial Overview: Challenges when training GANs Heuristics for Training Stable GANs Architecture for guidelines…
Read more
Highlight: In this post, we will briefly review the theory behind Generative Adversarial Networks and then we will learn to implement that knowledge in PyTorch. We will actually build our first GAN from scratch so that all the details are demystified. Initially, we will start with generator modeling or faking a simple 1D function (sine wave). In the later posts, we will build on the fundamental GAN architecture presented in this post. Tutorial Overview: GAN…
Read more
Highlights: Hello and welcome back. In this post, we will see how we can use Neural Networks for the segmentation task. To be more precise, it will be about Semantic Segmentation. The goal of Semantic Segmentation is to label each pixel of an image with a corresponding class. We will cover some of the most popular Deep Learning models for segmentation: Fully Convolutional Neural Network SegNet U-Net Tutorial Overview: What is Semantic Segementation? Types of segmentations Fully Convolutional Neural…
Read more
Highlights: Face detection and alignment are correlated problems. Change in various poses, illuminations, and occlusions in unrestrained environments can make these problems even more challenging. In this tutorial, we will study how deep learning approaches can be great performing solutions for these two problems. We will study a deep cascaded multi-task framework proposed by Kaipeng Zhang [1] et al. that predicts face and landmark location in a coarse-to-fine manner. So let’s get started! Tutorial Overview:…
Read more
