Jenny Han, Bradley Zhou, Jade Wang. Carnegie Mellon ECE Capstone, Spring 2021
This week, I did more in depth research on specific implementation details and design specifications. I revisited some of the previous decisions we had made, such as what Arduino modules for our pen and what libraries we would use for the gesture recognition. Previously, we …
This week, I researched the Arduino Pro Micro and HC-05 Bluetooth module compatibility. I looked into which pins should be connected to connect the two parts. I also created a Google Sheets to document our planned budget online. Additionally, I met with Jenny and Bradley. …
This week, the most significant risks that could jeopardize our success is ordering the correct parts for our project. Since so much of our project relies on the physical assembly and connection of parts that we have to buy, both time and money could be waste if we purchase the wrong ones. As such, we will do even more extensive research and be sure to create a budget outline to allocate for extra parts in case something does go wrong during our initial build.
As of now, since are have only begun our project, we’ve only just created a more so detailed design of our system. During this process, we did add some additional features for customization of body expressions after realizing that there is a lot we can do with Open Pose, the library that we plan on using to recognize and manipulate body expression inputs. This change was necessary because we wanted our project to be interesting and challenging enough given our manpower and time constraint for this semester long project.
Take a look at our final Proposal Presentation page!
This week, I did some research in ways that we could implement our project idea, as well as met with Jenny and Bradley to discuss the details of our project design. Last week, we had already solidified that our overall system would include a pen, …
This week, I did research on what kind of system we could build for our Capstone project.
For the pen, I researched different ways to measure position and relay information with the idea that the pen should be a standalone handheld device. I found some resources online and deciphered the best method based on size and compatibility with each other.
For body expressions, I looked into different ways we could track body positions. After doing some research, I think we could use OpenPose to track body landmarks. I found that OpenPose is best used with Nvidia GPU, so I debated the cost/benefit of allowing users to run the entire system on their personal laptop or having OpenPose be run remotely. I decided on running that remotely to depend less on the user’s computer.
In terms of parts, we researched what the average cost of components were. These components included GPUs, Arduinos, projectors, gyroscopes, etc.
In terms of designing the system, I tried to ideate as much as possible how many different ways we could have the user interact with our product and be very open with new ideas. From the ideation, we were able to come up with some problems and solutions.
Finally, I worked on the presentation. We had created a draft of notes we gathered from our research and I put it together into a coherent presentation that was visually pleasing and informative with as few words as possible. I added speaker notes for our presenter to use in case needed during the presentation. Also, I helped to create the Gantt chart for our schedule.
I think we are on time with the project so far. Our next step is to get started with OpenPose and order parts for the hardware elements. I am currently based in California (but I plan to fly back to Pittsburgh for this project), so I plan on installing and running OpenPose this week and getting the body expressions captured.
Draft Presentation with Notes: Proposal Presentation Draft
Final Presentation: Proposal Final Presentation
We will be creating Magic Mice, a laptop projector that can display any screen, anywhere. This attachment will connect to the user’s laptop (MacOS) and project the given screen onto any indoor wall. The user can use a pen device or body expressions to interact …