Segmentation methods for time-frames dataset

I am learning image segmentation meanwhile, and I can say that I am just a beginner, and I know some models used for segmentation (e.g. Unet). But lastly, I was working with a dataset where images are time-correlated, so the dataset is a time-frames taken from a video. So my question is: is it because we have this time correlation, some architecture models are better than others for this type of image? what do you recommend in such a case?
Thank you