The problem Zorohack solves

A person who is trying to stop using drugs can make mistakes, feel bad and start using again. This return to addiction is called a relapse. Studies reflect that 40% - 60% of individuals relapse within 30 days of de-addiction, and upto 85 in a year. It is important for individuals who struggle with addiction to acknowledge the higher risk of relapse and have an awareness of what their triggers are. In the present situation where the pandemic causes a lot of citizens to stay at home, people have spent a lot of time browsing and are likely to find triggers that cause them to relapse. Zorohack solve this problem by censoring each and every word and image which may lead to such a trigger.

Mainly our project does 2 things simultaneously:-
Censoring the words: Related texts are compared with the list of texts defined in the code and is censored.
Censoring the images: The images from the web pages are fed to our Machine learning model and predicts whether a particular image is triggering or not . To discard the triggers we alert the user and blurs the whole webpage.

Challenges we ran into

We had faced time constrain while making our project. We also faced difficulty while making the model to predict the images for our extension. Initially we did not know how to make a machine learning model on javascript .We overcame this problem with the help of google teachable machine. We also used a large number of dataset in our model to subdue the constraints faced by image recognition.