Release v1.1.1
Updated from 1.1.0 to 1.1.1 to hot patch a crash that occurred when changing confidence before selecting model.
Massive improvements to AI detection speeds, error catching (for help debugging), UI changes, and code improvements.
AI model no longer loads on UI load to help debug.
Resource usage has now dropped on GPU by a lot and CPU usage has gone up about double. For me this means instead of 5% CPU and 30% GPU I now see 10% CPU and 11% GPU.
AMD devices now run a lot better and don't have as bad of "choppy" frames.
No AI model improvements have been made with this update, hopefully next update we'll be retraining the model with 2-4k more images.