Improving genlock with AI (foveated video)

Before I try to get foveated 3D VR180 video panning inside 3D 360, I thought I'd try to get better genlock with AI interpolation. Genlock in 3D VR180 cameras is expensive because if this wasn't an issue, we could make affordable DIY rigs with second hand professional cameras such as bottom to bottom APS-C and fisheye lenses. Many of these may only film at 30 fps and achieving genlock is hit or miss but genlock is crucial. With AI interpolation, we can increase framerate and make it more easy to genlock the video, particularly if we own a high performance computer. The downside is that currently it's hard to do this often, even expensive high performance computers are slow to interpolate with AI and a more high end computer is required for example for foveated 8K60. It would be great if future AI software could be made that would not only interpolate the frames but also compare the left and right frames to match the images on its own and discard the rest of the frames so that you can input 2 non-genlocked left and right videos and output a genlocked 3D VR180 video.
I've interpolated this 4x to 4k88 on one phone and the other I've interpolated 2x to 4K60 and did my best to genlock them, it is easier than when I only had 4K22 (weak graphics driver in a custom ROM) and 4K30. Interpolation could be done to 4K120, 4K240 etc.
You can find the guide for adding text with Shotcut on the Deo VR blog: https://deovr.com/blog/89-shotcut-tutorial-adding-text-to-vr-videos-and-more