2nd Blog

Recap of Week Three: April 9th, 2021, 2021 - April 15th, 2021

Everyone's Progress


CJ Lin:

Attended meetings, worked on the pitch deck, and worked on writing the product requirement documents. In just two days, CJ rapidly prototyped simulating moving around in a virtual space and shooting a basketball into a hoop. Check it out!

Milestone for Week 3: Work Peter and Yan to understand what data can be used for feedback

Peter Michael:

Attended meetings, worked on the pitch deck, and worked on writing the product requirement documents.

Milestone for Week 3: Got pose estimation to run locally

Yan Zhe Ong:

Attended meetings, worked on the pitch deck, and worked on writing the product requirement documents.

Milestone for Week 3: Researched physics engines to used

Jasmine Woon:

Attended meetings, worked on the pitch deck, blog, and worked on writing the product requirement documents.

Milestone for Week 3: Finish up mock up/design of the application


Updates on Code

In two days we created a prototype that allows the user to move around in a virtual space, shoot a basektball in the hoop, and the ball will automatically return to the player. Check it out!

Peter has set up a dataset for pose estimation on his local computer

As of now, we are working on our individual parts. In the future, we will come together to make our projects more cohesive.


Updates on Ideas

Last Thursday, we had a Zoom breakout session with the VR course staff and pitched our idea for creating a card game idea. We were told to re-look at other proposals because our other ideas took advantage and heavily used the idea of being in augmented reality. After several email exchanges and meeting with our TAs, we decided to make an application to helps users better their skills at basketball shooting.

Our application will work by having the user place an external camera looking at the user. This external camera will track the player's movements. The user will place a virtual hoop onto their wall and throw a basketball into the hoop. After the player has thrown the ball, a retargeted model of how the player has thrown will be displayed. Using machine learning for pose estimation will determine how well the player has thrown. In addition, a textbox will appear to instruct the user on how to better their aim.

Check out our Pitch Deck and PRD


Plan for Next Week

Here are a couple of milestones that everyone plans on doing

Jasmine: Follow mock up design and place models in an environment

Peter: Do the networking with the AR Headset

Yan: Understand the chosen physics library

CJ: Measure key points and analyze data from the team to develop a feedback algorithm


Blocking Issues

One concern that we have is detereming the best approach in creating a feedback algorithm. We hope to find an existing algorithm that can help measure keep points of a person and from that data we can determine basic feedback like the user is shooting too far in one direction.