In this liveProject, you’ll take on the role of a computer vision engineer creating a proof-of-concept for a mobile app with world-changing potential. The goal of the app is to capture a picture of a skin mole, and then calculate the likelihood of it being a cancerous melanoma in need of medical intervention. Your mission is to build a model that can perform this image classification using low-resolution photos from a phone’s camera. You are provided with a training dataset, however, only a small fraction of the images have been labeled. To overcome this challenge, you will employ data augmentation techniques, transfer learning and semi-supervised learning with GANs to produce your final model.
This project is designed for learning purposes and is not a complete, production-ready application or solution.
This liveProject is for intermediate Python programmers with some deep learning experience. Knowledge of PyTorch will be helpful, but is not required. No prior experience in generative modeling, including GANs, is assumed. To begin this liveProject, you will need to be familiar with:
- Basics of PIL
- Basics of Matplotlib
- Intermediate Python
- Intermediate NumPy
- Classification as a machine learning task
- Intermediate deep learning concepts such as convolutional neural networks
you will learn
In this liveProject, you will learn important deep learning tools and techniques for data augmentation, semi-supervised learning, and generative modeling. These techniques are highly transferable to a wide range of machine learning roles, especially in the field of computer vision.
- Data augmentation built into the image pre-processing pipeline
- Pytorch and deep learning on the GPU
- Image classification using deep convolutional neural networks
- Transfer learning
- Generative modeling: variational autoencoders (VAE), generative adversarial networks (GANs)
- Semi-supervised learning with GANs for image classification