Storing "training" to use on subsequent videos

[John M.]

Hello all,

I have a question about using idtrackerai to track between videos. I'm using the GUI, no-GPU version on a PC, tracking a 1-min video with two fish, separated by a barrier. When I track one individual or the other, the entire process takes about 1.5 minutes. However, when I track both at the same time, it takes closer to an hour. I was wondering if there's a way to train the program on one video, but use that stored "training" of the fish shape to track another video (shot in the same style), so that the program doesn't have to go through the training sequence again with subsequent videos, which I assume is what takes up much of the hour that it currently takes to track multiple animals. I'd appreciate any insights you have on this!

Best,

John