odl_user_171120@udacitylabs.onmicrosoft.com rnmm66MCG*Hg
Ofurufu
means sky/aerospace in Yoruba, as in oko ofurufu
which means aeroplane.
Ofurufu, as a project, builds an automated passenger boarding system which comprises
- Passenger verification (Name & identity)
- Luggage validation: Flagging luggages with banned items (lighters, currently)
- Flight information verification/cross-referencing
Passengers are co-ordinated according to the verification results of this system. Additionally, the experience of the travelers are monitored using sentiment and emotion analysis.
All these are done with the aim of speeding up the passenger boarding time and improving experience for them.
This project relies on a number of Azure services and their SDKs
- Storage Account
- Face Cognitive Service
- Form Recognizer Cognitive Service
- Computer Vision Cognitive Service
- Azure Video Analyzer/Indexer
An azure subscription is thus necessary to setup and run this project.
- Clone this repository
git clone https://github.com/theyorubayesian/ofurufu.git
- Create a conda environment using the
environment.yml
file and activate it
conda env create -f environment.yml
conda activate ofurufu
- Create the above listed Azure services. Detailed instructions can be found here
- Upload the files in the
material_preparation_step
folder to Azure blob storage. This section should contain all the student deliverables for this project. - Train a custom Form Recognizer model for boarding passes using this Azure website. Use the boarding passes in the
material_preparation_step/boarding_pass/training_data
folder as training data. - Train a custom lighter detection model using the data in
starter/lighter_images
and Azure Custom Vision
All scripts are contained in the ofurufu while screenshots are in the relevant step folders (step_*/)