The problem FitMe solves

With increasing cases of COVID 19, the world has come to a halt. Public places have been shut to avoid large gatherings. One such place is Gym, which has made it difficult for so many people to keep up with their fitness. Due to the unfortunate position of isolation, people are struggling with various health-related problems such as obesity, irregular sleep patterns, eye strain, mental stress, decreased immunity, and hence, are at a higher risk of getting infected with the Coronavirus. The impact of staying fit is huge and it helps to be sane with all the stress people have. Now as the gyms are closed and people don't have the required guidance, an alternative is needed. In this pursuit of adapting to modern norms, a solution with gamified concepts is needed to help people with their fitness routine and keep them motivated to exercise daily.

Challenges we ran into

The biggest challenge we faced was to make the application both robust and lag free, as the Openpose model requires decent amount of computation which makes it heavy. To overcome this, we switched to another lighter model called PoseNet, which is an offering of Tensorflow. Another strange problem we encountered was that as the images of the subsequent poses were being shown to the user, the application would start misbehaving to the extent that the browser would freeze. We tried many different strategies to solve this problem and eventually found a satisfactory alternative of doing the same.