Description as a Tweet:

Enhance the appeal of your otherwise ordinary presentations by adding a touch of intelligent automation. Control your presentations on the go as you like! Play around with your presentation like a pro and provide an immersive experience to your viewers.

Inspiration:

Slideshows and presentations have dominated the education and business ecosystem for quite some time now. Being students ourselves, we understand the difficulties of working with a slideshow manually which range from mismanaged controls to even requiring an assistant. Our research surrounding this issue has shown that modern day enterprises and educational institutions struggle with control and flow of slideshows. There is an outpouring demand for improvements in this arena. Our project is a hack to solve this seemingly simple yet deeply rooted problem in a domain that is essentially pervasive in modern times. This project is our vision for a technology revolution - one that unleashes the vast potential and utility of automation.

What it does:

Our project is an easy-to-use GUI based tool that uses the principles of Natural Language Processing along with the powerful features of python to enable users to customize their slideshows. The tool empowers users to control slideshows in real-time through speech. The tool intelligently takes action by analyzing slide texts. This makes the implementation of the slideshow independent of any external influences and the user can present highly synchronized presentations just by speaking (well, literally!)

How we built it:

Code, Coffee & Errors!
We used our knowledge of python programming and the help of some popular python libraries to code the backend of our application using Visual Studio Code. As for the frontend, we experimented with python’s PyQT5 library to create an interactive GUI for our application. Finally, we used bootstrap and some HTML-CSS to build a web page for our application.

Technologies we used:

• HTML/CSS
• Python
• AI/Machine Learning

Challenges we ran into:

None of us have worked with the natural language processing in python beforehand. Similarly, we did not know how to build interfaces in python. We decided to extend the functionality of an already well built application (Microsoft Powerpoint) which is not only an unconventional approach but also quite challenging. We had to research and extensively use the documentation of the libraries we used to be able to write the code. To implement real time speech recognition, we had to use processing in parallel. None of us were aware of how to implement multiprocessing in python. None of us have previously been to a hackathon and hence, we were kind of unaware of what to expect. Managing time was also one of the major challenges.

Accomplishments we're proud of:

Through the course of this hackathon, we learnt the use of some new python libraries and integrations which will go a long way in helping us become better programmers. The fact that we were able to document our imagination and ideas into code successfully and complete this project in time is probably a big accomplishment for all of us.

What we've learned:

On the technical front, we learned new python libraries such as python-pptx, ntlk, PyQT5, etc. We also got an exposure to speech recognition and natural language processing. More generally, we learned skills of paramount importance such as time management, teamwork and collaboration.

What's next:

We believe that the strength of our project lies in its scalability. Our algorithm, though sufficient for now, can definitely be improved possibly by integrating some more complex machine learning models. We hope to add speech prediction (just like text prediction) to make the transitions more accurately. The application itself can be extended to perform other tasks that revolve around natural language processing (text summarizing, translating among others).

Built with:

We built our application in Python using python-pptx, nltk, PyQT5, a bit of CSS and passion.

Prizes we're going for:

• Best Documentation
• Best Venture Pitch
• Best Domain Name
• Best Machine Learning Hack
• Best Beginner Software Hack

Manan Talwar
Niharikka Tyagi
Manmeet Tuteja

Table TBD