FocusHacks

GIF
Focus Hacks Thumbnail
Image of all our classes on TensorFlow
I am focused in this capture, hence the value of focused is one.
I am taking a picture(invading privacy in a virtual meeting), hence violating privacy value is one.
I am writing, hence the value of focused is one.
I am using my phone, hence the value of unfocused is one.
Warning email when student takes photo (invading privacy)
Email that gets sent when user loses focus (when Focus index = 0)

Inspiration

Studies show that approximately 80%–95% of university students engage in procrastination to some degree, and about half procrastinate in a consistent and problematic manner[1]. 50% of students do not get to make the most out of their day because they get distracted, and FocusHacks is here to change all of that. This web application uses Machine Learning to visually track when the user loses focus, and sends a reminder/motivational quote through email to get back on track, and make the best out of their day.

What it does

1. Enhances user’s productivity

FocusHacks helps users reduce distractions and increase productivity.
With its machine learning model, FocusHacks could detect when the user is being productive and when he/she is not.
If the user gets distracted and is not focusing on his/her work, FocusHacks would send the user an email, reminding and motivating the user to get back to work.

2. Protects online privacy

The technology of FocusHacks could be further applied in virtual classrooms.
FocusHacks detects when students are trying to take photos of the virtual classroom with his/her cellphone.
If the student is violating the privacy policy during remote learning, warning emails would be sent to both the student and the teacher.

How we built it

Machine Learning model: Tensorflow
Frontend: HTML5, CSS3 and JavaScript
Backend: Node.js, Express.js
API service for sending emails: Twilio SendGrid

Challenges we ran into

Machine Learning model needing more training data
Collaborating with different time zones
Wifi cutting out accidentally

Accomplishments that we're proud of

Built a Machine Learning model that could classifies user's behaviours, and detects users’ behaviours when they are not focusing on their work or violating the privacy policy
Capable of sending email reminders to users promptly when the Machine Learning model has detected the users not being focus on their work

What we learned

Collaborating with people with different backgrounds, skills and experiences
Using Git for collaboration and version control
Building a Tensorflow model
Using Twilio's API to send emails to users
Building a web app with both frontend and backend technology

What's next for FocusHacks

Using Auth0 to register and authenticate users to use our service
Creating a more powerful machine learning model that could identify more classes
Creating a mobile app where you can see your performance data
Using Twilio’s communication API’s to far greater levels for user communication to be even more effective
Storing our data in database such as PostgreSQL, MongoDB or Firebase

Citations

[1] Steel, P. (2007). The nature of procrastination: A meta-analytic and theoretical review of quintessential self-regulatory failure. Psychological Bulletin, 133(1), 65–94. https://doi.org/10.1037/0033-2909.133.1.65

Built With

Submitted to

NewHacks 2021
- Winner Most Creative Use of Twilio

Created by

I formed the various animation concepts within the video. I also helped increment the design of the site.

Danial Raheem
I trained the Machine Learning Model using thousands of images exhibiting the different behaviors. I was the primary tester for debugging the Twilio API with the Machine Learning Model.

Mustafa Abdulrahman
Software Developer - Communication Enthusiast
I worked on the frontend and backend development. I also worked on the integration of the Twilio API into the project.

HSKPeter Hui
I Worked on creating and training the Machine learning model with TensorFlow, Integrating the TF model into the webpage, Coded the front end of the webpage, Tested and debugged The ML model along with the Twilio API, Edited, screen recorded and made the voiceovers for the YouTube Video.

Ibrahim Kasim
CS @ UofT