What data used in training?

[yw0nam]

Hi, i want to ask about data for training.
In, paper, LRS2, LRS3 are used for training and validation.
So, I download LRS2, LRS3 dataset.

These datasets consist of videos that has cropped face from original video.
But, in inference, Input videos are original videos that has not cropped face.

So, I want to ask whether you used cropped video or original video when training the model.

Thanks.

[fake-warrior8]

are the LRS2 and LRS3 datasets pre-training dataset? I can not found any other description about the pre-training dataset.