Platypus | Devpost

Logo

Inspiration

*An ML model is only as good as its training data. * Data scientists know that an untrained statistical model is next to useless. Without high-quality labeled training data, supervised learning falls apart and there is no way to ensure that models can predict, classify, or otherwise analyze the phenomenon of interest with any accuracy.

What it does

Human cognition is a boundless resource on the Internet of you’re clever enough to leverage it for labeling tasks.Our app helps in outsourcing the task of labeling to crowd-oriented environments. We have created an Android app which helps in presenting training data in simple manner to help users identify, classify, or otherwise comment on images, text, objects, and other presented entities. Also an additional matrix that we built in was using Azures Cognitive vision API to extract the contents of an image and verify them to give a user metric to ensure quality of the labelled data set produced

How we built it

We have used Android studio as our primary IDE along with Azure Cognitive service to built our app.

Challenges we ran into

None of had any expertise in building mobile apps. We took it as a challenge to learn app development from scratch and create an MVP in the small time frame

Accomplishments that we're proud of

A working mobile app with basic functionalities

What we learned

We learnt how big of a problem it is in machine learning to get labelled data sets being Big Data Analytics Majors Having experienced it first hand. Also technically we ran into a number of problems from setting up android studio to making the swipes work and annotating over the image