Automated shopping using Neural networks

Use computer vision to automate billing

Project

The project comprises of three modules:

1. Object Detection
2. Pose Estimation
3. Activity Detection

Directory structure

.
├── data
    ├── demo                        # contains demo files
    ├── frames                      # raw images
    ├── labels                      # object detection annotation data
    ├── mod-data                    # activity detection annotation data
    ├── images                      # directory for output images
    └── persondata                  # pose estimation annotation data
├── project    
    ├── main.py                     # the system code
    ├── activity_detection          # activity detection module code
    ├── object_detection            # object detection module code
    └── pose_estimation             # pose estimation module code
    ├── slim                        # slim directory to set up object detection module
    └── setup.py                    # code to set up object detection module
├── tools                           # annotation tools
├── requirements.txt                # pip freeze of anaconda environment
├── environment.yml                 # the exported anaconda environment file
├── Dockerfile
└── README.md         

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.

Prerequisites

• You need to be on Ubuntu - as Tensorflow for Python 2.7 isn't available for Windows :(

Installing

1. Using Anaconda

• The packages are installed using Anaconda, so you need to have them installed first. Anaconda installer can be downloaded from here

• Clone the repository. You can do this by entering the following command in the terminal window. To open a terminal window, either right-click and select Open terminal or press Ctrl+Alt+Enter.

  $ git clone git@github.com:nabhanabdulla/MLabs.git

• Change directory to the clone folder

  $ cd MLabs

• Now, type this command to create a new anaconda environment with the dependencies for this repo

  $ conda env create -f environment.yml

• Activate the environment

  $ conda activate MLabs-cpu

• Download the pose estimation model from here.

• Now, type this command to go to directory where the downloaded model should be placed.

  $ cd project/pose_estimation/model

• Place the downloaded model in the current directory.

• To try running the project, goto the projects folder

  $ cd ../..
  $ pwd
  /home/username/**some-directories**/MLabs/project
  $ python main.py

Authors

• Karthik - automated_shopping_experience
• Hemanth
• Abin

Acknowledgments

• MLabs