My project is title Bridge, it is an AI powered sign language interpreter, and this week I have started programming the software. In this Journal, I will talk about the progress and limitations of the project.

So far, here are the progresses of my study:

• Research of various methods to optimize my model training is going well and have a lot of researches possible
• I have started programming the data collection part of my program and intend to finish by this week to start training
• I have borrowed high-definition web cameras from the IT and will make the data collection process easier

Able to use IMA computers for training

This week spent a lot of time trying to figure out how to use the high-performance computers available in school in order to have a more efficient training process.

I first tried to use the Mac’s and pc’s at the library however I couldn’t add the necessary libraries for the software, it is mainly because I do not have admin privileges and thus cannot add things to the local drive

Then I tried to use the IMA computers in school, they worked! I have been using them and have some progress in the process of programming.

A problem with this is that I cannot use the computers when in the dorms, I have been asking around and have discovered that the school offers the use of virtual machines in the server and the use of HPC for research with the permission of the research adviser.

I also tried to figure out how to use the dorm macs to program. I have succeeded by using things such as virtual environments in which I can localize the libraries that I want to use. However, this has a lot of bugs which I cannot fix as of now

This had cost me a lot of time just to figure out, however I think that being able to use virtual environments is a win as it will make the packaging into into an app much easier in the long run

Making the UI

Through the weekend, I have tried to research about the different ways to create UI’s for python programs, in this process I have discovered the app called PyQt6, where I can do some drag and drop functions. I have watched a full course about it in Saturday and tried to implement it into my own code. This adding the UI process is actually really difficult and has proven to be a problem. I am forced to use OOP in programming and thus have to refactor most of the code that I have already made.

Code converted to OOP for GUI(left), Original code on right

So far, I have only tried to put the box on the left side of the screen with some transparency for the live transcription, I tried to mix this with the OpenCV camera and the media pipe detections and have seen a drastic amount of dropped frames, everything becomes laggy and virtually unusable. I have yet to find the solution for this.