Is there any technique for animating just the face in a larger source image? That is, the driving video would be 256x256, the face in the source image would fit in a 256x256 space but the entire source image would be bigger? See attached image where the face is a 256x256 area in a larger image.
I could extract the face into a 256x256 image, do the animation and then composite the output video and original image in a video editing package. But it would be nice if there was a way for the algorithm to find just the keypoints for the face in a larger image and animate just the face area directly.
Is there any technique for animating just the face in a larger source image? That is, the driving video would be 256x256, the face in the source image would fit in a 256x256 space but the entire source image would be bigger? See attached image where the face is a 256x256 area in a larger image.
I could extract the face into a 256x256 image, do the animation and then composite the output video and original image in a video editing package. But it would be nice if there was a way for the algorithm to find just the keypoints for the face in a larger image and animate just the face area directly.