HusseinLezzaik / Gestures-Classification

Classification of Human Gestures using different models like AlexNet, VGG16, ResNet pre-trained on ImageNet

Gestures Classification

Introduction

Classification of Human Gestures using an in-house dataset collected from the students of my class, using different models like AlexNet, VGG16, ResNet pre-trained on ImageNet using PyTorch and Transfer Learning.

Overview of the Repository

In this repo, you'll find :

• dataset: pictures of my classmates thumbs labelled up or down.
• cat: pictures of different kinds of cats to test pre-trained VGG on ImageNet.
• imagenet_labels.py: ImageNet labels for 1000 class.
• like_classifier.py: thumbs classification for thumbs up/down using VGG-16.
• VGG_classifier_cats.py: cats classifier using VGG-16.

Getting Started

1. Clone repo: git clone https://github.com/HusseinLezzaik/Gestures-Classification.git
2. Install dependencies:

   conda create -n gestures-classification python=3.8
   conda activate gestures-classification
   pip install -r requirements.txt
   

3. Run like_classifier.py for classifying thumbs as up/down, you can play with the test set by yourself ;)
4. Run VGG_classifier_cats.py to discover the different kinds of cats on our planet.

Contact

• Hussein Lezzaik : hussein dot lezzaik at gmail dot com