CarSpottr is a web application that allows users to upload images of vehicles and get 3 segmentations of the images, corresponding to three different trained models.
This app is part of my final degree project, which aims to answer of whether it is possible to train a model to segment images of vehicles by using images generated from 3D renders as a form of data augmentation.
I used DeepLabV3+ as the model for the segmentation, and trained it using the a combination of datastes: one soruced online, and one generated from 3D renders.
- Get Docker Images
docker pull lumialfe/carspottr-webapp
docker pull lumialfe/carspottr-restapi
- Start the project
docker compose up --build
- Shut down the project
docker compose down