chenhuafan / 641-541-final-project

541 Final Project - Gesture Recognition

641 Final project

Author:

• Bhumi Godiwala ([email protected])
• Chenhua Fan ([email protected])
• Qiwei Chen ([email protected])

Quick Start

install the requirements.txt by pip, then go through the *.ipynb.

Or using mlflow to start mlops UI. mlflow server --backend-store-uri sqlite:///641.db --host 0.0.0.0 --port 5000 SQLite required.

Files

All necessary files can accesse at the google drive.

This repository is the main workspace of our final project. It includes:

• /GAN is the main GAN model we used and trained.
• *.ipynb Jupyter notebooks of model we used (main work).
• ASLDataset.py the initialization of dataset.
• /trained_models the models we trained (Models not included in the zip file on the Canvas).
• /imgs_outputs is the images generated from GAN.

Models details

All models not included in the zip file on the Canvas. Additional upload required.

• model ResNet model.
• ST-CGAN_G1_1260.pth GAN model. 111.6MB
• ST-CGAN_G2_1260.pth GAN model. 111.6MB

MLOPS

the mlops based on mlflow. data stored in the 641.db (sqlite)