Unscripted


  • Machine Learning


Mentors :

  • Anirudh Mittal

  • Siddesh Pawar

  • Pranav Jeevan

Mentees :

  • 4 students


Description: Journalists usually go out with their microphone to various places speaking into the recording device to note important events. They use these recordings later to prepare their article. We envision a system that can take a few minutes of audio from a recording device and use that information to write an article on that event. This fundamental concept can be extended to many other applications one of which can be automatic minutes-of-the-meet taker.

We’re looking for team members who have experience with programming. Experience with ML, more specifically speech recognition and NLP will be a plus. Some articles to begin with: https://medium.com/@ageitgey/machine-learning-is-fun-80ea3ec3c471
https://medium.com/@sanchittanwar75/introduction-to-machine-learning-and-deep-learning-bd25b792e488
https://www.youtube.com/watch?v=aircAruvnKk


Tentative Timeline :

Week Number Tasks to be Completed
Week 1 Basics of machine learning and webscraping
Week 2 Basics of Speech-to-text, NLP and APIs
Week 3-4 Data collection and preprocessing
Week 5 Traning the model
Week 6 Finetuning the model
Week 7-8 Implementation

Checkpoints :

Checkpoint Number Progress
1 ML Basics
2 Domain specific knowledge
3 Cleaned data
4 Model
5 Final product