INNER CODE UNIT · Python

source_image_face_emb

jdh-algo/JoyHallo · scripts/app.py:329

    source_image_face_emb = source_image_face_emb.reshape(1, -1)
    source_image_face_emb = torch.tensor(source_image_face_emb)

    source_image_full_mask = [
        (mask.repeat(clip_length, 1))
        for mask in source_image_full_mask
    ]
    source_image_face_mask = [
        (mask.repeat(clip_length, 1))
        for mask in source_image_face_mask
    ]
    source_image_lip_mask = [
        (mask.repeat(clip_length, 1))
        for mask in source_image_lip_mask
    ]

    times = audio_emb.shape[0] // clip_length
    tensor_result = []

View source record →