WiiGym

Inspiration

Working out with the friends you made at nwHacks can be difficult, especially when you all live a plane ride away from each other✈🌎. Along with that, it's hard to stay motivated when you're working out by yourself, with no good friendly competition. Michael, and I (Julian) both have recently started working out, with Michael hitting a personal PR on bench by repping 95 lbs 👏. But after nwHacks, we won't be able to see each other ever again, and our desire to workout will diminish!

But it doesn't have to be this way! By integrating competition with live video feed, we created an application that gears to tailoring our problem of demotivation and disconnection through friendly competition, whilst putting those reps in.

One study found that 95 percent of those who started a weight-loss program with friends completed the program, compared to a 76 percent completion rate for those who tackled the program alone. The friend group was also 42 percent more likely to maintain their weight loss.

Let's think about the real world friendships and connections that working out in real life bolsters. Creating physical gains with friends is the ultimate boost to intrinsic motivation.

Clawing our way through this problem, WiiGym (WeGym) integrates video calling with body weight exercises by combining AI tracking with multi-user video calling. We kept ease of repeatability in order to ensure users are working on their long term fitness goals, and integrated a leaderboard to see were you're at.

What it does

WiiGym offers an easy way to workout with friends. You can customize your workout experience by selecting accuracy, rep count, or a timed duration. Simply invite your friends, select whatever option works best for you, and start exercising! WiiGym will take in your exercising statistics and provide them through points over workouts, and a leaderboard with other users.

How we built it

Our application combines WebRTC and MediaPipe for users to see their "skeleton model" which stores each user's rep count in a MongoDB Atlas database. By integrating our improvised and combined WebRTC and MediaPipe model in Grok, our program became runnable on the local host.