Welcome to the repository for the CNN Tutorial using PyTorch. This repository is designed to guide participants through the fundamentals of Convolutional Neural Networks (CNNs), building models with PyTorch, and applying them to real-world data.
This repository is divided into the following sections:
- Prerequisites: Resources for getting used to Python and Google Colab.
- Demos: Step-by-step instructions and content for the five sessions.
- Exercises: Practical assignments to be completed during the session to reinforce the concepts learned in the demos.
- Solutions: Solutions for all the above Practical assignments.
Detailed instructions can be found in the prerequisite folder
| Demo | Handled By | Topics Covered | Exercise | Folder |
|---|---|---|---|---|
| Demo 0 Demo 1 |
Gowthamaan | Introduction to Pytorch: Tensors, Pytorch-Workflow, Basic CNN operations, Building a CNN model, Classification (FasionMNIST) | Exercise 0 Exercise 1 |
D0 Folder E0 Folder D1 Folder E1 Folder |
| Demo 2 | Sidharth | Object Detection, SSD | Exercise 2 | D2 Folder E2 Folder |
| Demo 3 | Abhikesh | Segmentation, Transfer Learning | Exercise 3 | D3 Folder E3 Folder |
| Demo 4 | Kunal Patil | CNN Applications - ASL Letters Detection, Object Detection | No Exercise | D4 Folder |
-
Bourke, D. (2024). pytorch-deep-learning: Materials for the Learn PyTorch for Deep Learning: Zero to Mastery course [GitHub repository]. GitHub. https://www.learnpytorch.io/
-
Wang, Z. J., Turko, R., Shaikh, O., Park, H., Das, N., Hohman, F., Kahng, M., & Chau, D. H. (2020). CNN Explainer: Learning Convolutional Neural Networks with Interactive Visualization. IEEE Transactions on Visualization and Computer Graphics (TVCG). IEEE. https://poloclub.github.io/cnn-explainer/
We have drawn inspiration, code snippets, and implementation details from different sources for the exercises and demonstrations included in this repository. The resources we have utilized are duly cited in the references section. We have made every effort to ensure that all contributions and sources are properly acknowledged. However, if we have inadvertently missed any source or reference, please feel free to bring it to our attention, and we will update the repository to reflect the necessary credits.