Inspiration
Our inspiration came from common discovery channels and nature documentaries we all watched as children. We reminisced about the times we had marvelled the knowledge the people exploring the wild had about it. They knew exactly what to do just by seeing it, they were able to identify what was dangerous and what was friendly in nature. This knowledge encouraged us to explore how we could replicate such knowledge at our finger tips. That is why we have created the "Discovery Lense" that gives you the knowledge of any expert regarding any matter, given the right data set.
What it does
Using training data about any animal it has the capability to identify it. Currently, due to limited time and resources it is only able to identify objects in the real world like vehicles and domesticated animals. However, if we change the training set it can identify anything from species to even furniture prices. All the user must do, is plugin the right data set and watch the magic happen. It can even perform in real time identifying objects as they move.
How we built it
We built this using google collaborate and pytorch which enabled us to use preloaded data sets and numerous functions making it significantly faster and efficient.
Challenges we ran into
A challenge we ran into was debugging the training set as it was unable to run on our mac laptops, we eventually upon doing some research found out that the ".cuda()" command is unable to run on anything without an NVIDIA driver. Another problem we ran into was procuring a data set that matched our needs but if this product gets popularized it is capable of incredible things.
Accomplishments that we're proud of
This very project is the accomplishment we are proud of, it iis something capable of revolutionizing the modern world and it already has with numerous applications like the google lense and even in the security cameras placed on traffic lights.
What we learned
We learned how to collaborate and to work as a cohesive unit to accomplish a greater goal. Me and my team-mates were inspired by this hackathon to create a change in the world of education, this simple mechanism can make the world a better and efficient place. It encouraged us to use critical thinking and problem solving skills to solve any problems that arise with the idea and with the code itself.
What's next for Discovery Lens
We hope to move this to a bigger system and truly let it flourish with that NVIDIA graphics driver to see what it is capable of, possibly even finding better data sets to educate the youth and anyone wishing to simply learn more about the environment they're in. "Knowledge at your fingertips", that was the goal when the internet was first created, we hope to further this goal an to utilize these aspirations to create a better Earth.
Log in or sign up for Devpost to join the conversation.