Describe what you are trying to accomplish and why in non technical terms
I want to be able to have full types and boundaries visible in training images to be sent Frigate+ so that I can more easily and accurately train.
I have several cars that are always present on one of my more active cameras, so when I need to train an image from that camera, I have to accurately select the car and license plates, along with whatever I am really training for. Often two people, their faces, and a dog.
It would be excellent if all of that data could be sent to Frigate+ rather than just the one boundary and type that happens today.
Describe the solution you'd like
I think it would make sense to run a detection for all types and boundaries visible in the image locally before it is sent to Frigate+.
Describe alternatives you've considered
The Frigate+ server could run the detection, but that needs a bunch of computation and needs the model to be present on the server along with acceleration hardware.
Describe what you are trying to accomplish and why in non technical terms I want to be able to have full types and boundaries visible in training images to be sent Frigate+ so that I can more easily and accurately train.
I have several cars that are always present on one of my more active cameras, so when I need to train an image from that camera, I have to accurately select the car and license plates, along with whatever I am really training for. Often two people, their faces, and a dog.
It would be excellent if all of that data could be sent to Frigate+ rather than just the one boundary and type that happens today.
Describe the solution you'd like I think it would make sense to run a detection for all types and boundaries visible in the image locally before it is sent to Frigate+.
Describe alternatives you've considered The Frigate+ server could run the detection, but that needs a bunch of computation and needs the model to be present on the server along with acceleration hardware.