MLops

Tested:

Project Overview

This project is an implementation of a food classification system and pose estimation model.

For the food classification, we use an ensemble of EfficientNet, InceptionResNet, and ResNet models to achieve high accuracy.

For the pose estimation, we use MediaPipe, a cross-platform framework for building multimodal applied machine learning pipelines.

Backend Installation

Manual with docker:

docker build <server Name> .
docker run -p 5000:5000 <server Name>

Manual with virtualenv:

git clone git@github.com:ACT-HealthWatch/food_classification.git  
python3 -m venv <project name>  
rsync -av --exclude food_classification <project name>  
cd <project name>  
pip3 install -r requirements.txt  
gunicorn app:app