- Argo has been intergrated into AVP app
- Speech to text is working (still needs debugging)
- Issues when trying to start up a second recording
- Text to speech working
- Started looking into UE5 and Unity for VR development
- Looked into building 3D mesh of protein-protein data
- Unity requires a pro license for visionOS development, so it has been dropped.
- UE5 was refusing to build and reported the following error: InvalidDataException: The archive entry was compressed using an unsupported compression method. Fixed it by changing 'http' to 'https' in the BaseURL in Engine/Build/Commit.gitdeps.xml Source
- Got hand gestures working. Using hand gestures for basic commands
- Stores conversation with Argo, so we can have an on going conversation. (still needs work, continuesly mentions previous messages, even when not brought up)
- Successfully converted dot file to usdz and imported into Apple Reality Composer Pro.
- Started building interface for the string-db API
- Set up interface for Gecko, our protein visualization tool. We can now query the STRING-db with requests for data.
- Created functions that deal with meetings. It now transcribes, summarizes, and names them.
- Started working on functions to generate 3D models for protein data.
- Created features for meetings. Now keeps tracks, summarizes, and names meetings.
- Finished functions to create spheres for each protein returned by API request.
- Finished tooltip-style description boxes for protein objects that appear when clicked on.
- Added edges to protein network
- Able to connect to Polaris via sending HTTP request to a server
- Working on GENIUS animation
- Setup "show me" command to use Sketchfab API to search for and choose a 3D model to open using Argo for decision-making
- Made single gesture to use while recording instead of seperate gestures for start and stop
- Made a working terminal in the app
- Began working on integrating Argo with the terminal
- Made window for displaying simulation video
- Ran benchmarks to compare Argo/GPT 3.5 vs Llama 3
- Argo runs faster
- Llama response is easier to manipulate (can tell it how to start or end response)
- Running jobs on Polaris and displaying results
- Using LBM CFD sim
- Able to display the results as a video
- Created window to display simulation
- Made recording one held gesture instead of two for starting and stopping
- Error recovery and guiding the user
- Getting protein objects to move around 3D space.
- Getting edges to follow protein movement.
- Able to submit jobs from the app
- Integrated Argo into running and modifying jobs
- More UI changes: Fonts, color schemes
- Removing jitter from edge movement.
- Fixed small errors in running the simulation
- Integrated Argo to provide protein interaction partners.
- More UI changes
- Working on trying to convert a .vtu to a .usdz
- Adjusted simulation feature to allow adjustments of parameters before running the job
- Put a pause on trying to get a simulation that produces a 3D module working
- Working on making a calendar/scheduling feature
- Added clear all windows function