Attention detection

Build a Vision AI which detects an attentionate driver.

AI abilities :
● take an image as input.
● output the probability that the driver is paying attention to the road.


Suggested Dataset :
https://www.kaggle.com/competitions/state-farm-distracted-driver-detection/data



A notebook (html or ipynb) :
○ Data exploration
○ Models training
○ Performance evaluation

● BONUS: Script (link toward public repository on github/gitlab/bitbucket):
○ A quick application taking an image as input.
○ Returns the predicted probability of driver attention.


Introduction
Using at least 3 resources (kaggle notebooks, blogs,youtube videos or else), explain what methods exist to solve your problem and which one you decided to use.

Modelization
Explain the different steps you decided to implement. If you use a deep learning approach, please explain the architecture of your model and how it works. Don’t forget to cite the work of others.

Performances
● A baseline is defined.
● The performance of the data exploration choices are clearly compared and explained.
● Graphs are readable (size, density etc.) and understandable (titles, legends etc.)


Application (Bonus)
● BONUS: Script (link toward public repository on github/gitlab/bitbucket):
○ A quick application taking an image as input.
○ Returns the predicted probability of driver attention

● Take an image as input.
● Use your selected model to predict.
● The predicted result is clearly visible and understandable