Doesn't the Tiktok model creates a fake depth map from what the picture would be so if you fed it into a picture of a picture you'd actually just get a depth ma

Doesn't the Tiktok model creates a fake depth map from what the picture would be so if you fed it into a picture of a picture you'd actually just get a depth map of the original picture not a flat plane. You'd need to verify through the actual device camera sensors.