Pursuit for Opportunity: Pioneer November 2018

3 min readDec 21, 2018

This time, our idea was to make a smart survey form builder that uses computer vision to detect outline structure of logic jump style forms using computer vision and voice commands. A bit of an overkill but a cool project nevertheless.

Idea:

Let us say you want to make a survey form where each user is asked a different array of questions, depending upon the response they give. For example, if to the question of: Do you know programming? w have:

Yes -> Which Language are you familiar with?
No -> Would you like to?

So we have a branching at this question (logic jump) and this better helps us evaluate an individual. Based from my friend Bharat’s research work and Kurians innovation we thought what if we made an alternative where:

You draw using pen and paper the structure of your form using blocks to indicate questions and arrow marks to show possible routes.
Take the app on your phone and scan the paper to make sense of the symbols and automatically render the same template survey form.
Now using voice recongnition, we read out the questions and fill in the details for each. If it worked, this could take about 5 minutes to make even the hardest logic survey form as long as you can get it right on paper.

Though the idea was promising, unfortunately we faced few problems:

Typeform is excellent when it comes to branching forms. Why would users want to switch?
None of us were familiar with the required technologies (image recognition, d3.js etc) and so learning them amidst our hectic end semester was challenging.

So we could not finish it but plan to eventually do so. These were the responses we received from each week of the tournament: