Galp_CarWash_Report(1) is the second and final version of our report. We did a deep explanation of our solution. The content of this file is organized as follows: 

1) Delimitation about the problem; 

2) Hypothesis of Solution; 

3) Team Members;

4) Team Strategy;

5) Technologic pre-requisites;

6) Datasets;

7) How the solution can be used?;

8) Metrics and Results;

In the notebook folder, you can find all experiences that we tested in this competition. notebooks/dirt_extraction.ipynb file is the file that contains the deep Learning training model that extracts the features related to dirtiness and classifies if the vehicle is dirty or not.

The notebooks/car_detection.ipynb file contains the car detection model, which identifies the vehicle's precise location or concludes that an image doesn't include a car.

Finally, the notebooks/weather_classification.ipynb notebook is similar to the first one but tries to predict the weather based on an image. However, we didn't have time to add this part to the final solution.

We were not able to upload the images to Github because we would exceed the space limits (Github limitation). For that reason, you will not be able to run the notebooks.

To present the solution, we implement an application, and you can find the source code in this directory: src/

This main script uses two script auxiliaries. (src/utils/ and src/utils/

The final models used in the application are stored in models.

To run our application, you have two methods:

(1) Using our web application that is online:

(2) Using Docker

Run the following commands:

2.a) docker pull josemota/linear_team_app:latest

2.b) docker run -p 8501:8501 josemota/linear_team_app:latest