Inspiration
Many students struggle with their sciences. This is because scientific topics such as Chemistry, Biology, or Physics are typically complex and difficult to understand. It is for this reason that we decided to create SciPulse; To help students in their science endeavors.
What it does
SciPulse has a multitude of functions. Firstly, a student is able to upload an image of an app to the app in order to receive information on the animal. Students are also able to upload pictures of their chemistry titrations to receive feedback on their titrations. Additionally, students are able to ask the science wizard any problem, particularly physics problem to receive detailed help on their inquiries.
How we built it
SciPulse was built using Tensorflow to classify animals and titrations. GPT-3.5 Turbo API was also used to generate information on the animal once it has been classified. Python's Kivymd library was used to build the frontend of the app.
Challenges we ran into
The primary challenges which we ran into includes optimizing the accuracy of our Tensorflow models, connecting Kivymd to the backend of the application, and gathering a large enough dataset to train our Tensorflow models.
Accomplishments that we're proud of
We are proud of how quickly we were able to learn to use kivymd, tensorflow, and opencv and put it into one project.
What we learned
While developing SciPulse, we learned to use kivymd to create mobile applications, Tensorflow to create deep learning models, the troubles of overfitting your Tensorflow models, and the steps needed to use the OpenAI GPT API.
What's next for SciPulse
*********Making it more widely accessible. Other science topics.
Log in or sign up for Devpost to join the conversation.