Yoga-Pose-Estimation

A deep learning model that classifies yoga poses effectively using Convolutional Neural Network (CNN) by learning from a collection of input images.

This was a project done as part of Deep Learning for Computer Vision course at my university. The project was accomplished using 2 different deep learning methods:

1. Feed Forward Network
2. Convolutional Neural Network

So in this repository, you will be able see:

├── Project Source Code
|   ├── Phase_1_Feedforward[Source_code].ipynb
|   └── Phase_2_CNN_Efficientnet[Source_code].ipynb
|   └── Phase_2_models_for_comparison[other_than_efficientnet].ipynb
|   
├── Project Documentation
|   ├── Phase-1-Feedforward-report.pdf
|   └── Phase-2-CNN-Efficientnet-paper.pdf

Dataset

The model developed for the objective was trained using the ’Yoga Pose Image classification dataset’ from Kaggle. Data augmentation was performed on the dataset at later stages.

Kaggle Dataset: Link

Contributors

Anand K S

Rahan Manoj