D

Detectinator

Respire. Detect. Inspire.

D

Detectinator

Respire. Detect. Inspire.

The problem Detectinator solves

A project which can be used by soldiers to quickly check their respiratory health.

Be it the challenging altitudes of Siachen, or the deep waters of the Indian oceans, our soldiers brave them all to secure our nation. One constant problem that hinders their call of duty, is the lack of efficient and quick respiratory healthcare testing and screening. Moreover, due to the extreme conditions and lack of proper infrastructure, a lot of precious time is wasted in the transport, storage and maintenance of the testing kits, tools and samples. This is where our technology comes in. We have created a smart platform using AI which effectively detects and provides accurate results in respiratory diseases like the ongoing Covid-19, pneumonia and asthma. Soldiers will be able to easily access the platform and check for any such respiratory issues instantly.

Challenges we ran into

  1. The size of the datasets we had were small so we had to preprocess the dataset and make sure that our CNN model doesnt overfit.
    2.We also attempted to use GANs to create synthetic data to solve this problem.
    3.Cleaning the input taken at runtime proved taxing as we had to google multiple python libraries and code and recode the same functions multiple time so as to get the best possible result.
    4.While we were able to create a CLI-Based python program, translating that project to a web-based project proved hard as were not able to capture the voice input in the form of .wav files . We had to simultaneously try working on it via Django and Flask

Discussion