The problem Shadow Fighters VR solves

This is a game and we developed it for sheer fun. But while developing we realised pose mapping and interacting with it in Virtual, Augmented or Real world can be way too useful. In the sense, Virtual world collaboration in sports or cultural dance can be done in no time with just a cheap VR phone and a webcam. In augmented world this can be used in real time collaboration or learning something like self-defense, yoga or a normal work-out. In real world the pose can be mapped to machines. This can be helpful in controlling light or heavy machinery using human pose and gesture. This increases safety and stability in the field of construction, manufacturing or Defense. Refined-level pose estimation can do even more wonders. It can help the autonomous cars on the road to judge human behaviour and predict the next move. For example, predicting whether a human standing on the pavement will cross the street or not.
This technology if used properly can be Useful, Human-intuitive and Life-Saving.

Challenges we ran into

There were many challenges, but the work of a hacker is to either tackle them head on or bypass them. We did the same and there is nothing that interesting to talk about. But one interesting thing that we ran into is worth sharing. Pose estimation and mapping can be done in more than ways you can think and there are several ways to improve the tech depending on how you want to use it. Let's keep this as a surprise for our final product.