brige/human
1
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases 2 Wiki Activity

modify inter handler

Browse Source
This commit is contained in:
jiegeaiai 2024-11-20 22:05:39 +08:00
parent 1bfb79b45a
commit f86368bc37
3 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
data/avatars/wav2lip_avatar2/face_imgs/00000000.png
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 16 KiB

After

Width:  |  Height:  |  Size: 26 KiB

4
human/audio_inference_handler.py
View File

@ -59,7 +59,7 @@ class AudioInferenceHandler(AudioHandler):
super().on_message(message)
def __on_run(self):
wav2lip_path = os.path.join(current_file_path, '..', 'checkpoints', 'wav2lip_gan.pth')
wav2lip_path = os.path.join(current_file_path, '..', 'checkpoints', 'wav2lip.pth')
logger.info(f'AudioInferenceHandler init, path:{wav2lip_path}')
model = load_model(wav2lip_path)
logger.info("Model loaded")
@ -130,7 +130,7 @@ class AudioInferenceHandler(AudioHandler):
img_batch = torch.FloatTensor(np.transpose(img_batch, (0, 3, 1, 2))).to(device)
mel_batch = torch.FloatTensor(np.transpose(mel_batch, (0, 3, 1, 2))).to(device)
# print('img_batch:', img_batch.shape, 'mel_batch:', mel_batch.shape)
print('img_batch:', img_batch.shape, 'mel_batch:', mel_batch.shape)
with torch.no_grad():
pred = model(mel_batch, img_batch)

2
human/human_context.py
View File

@ -18,7 +18,7 @@ current_file_path = os.path.dirname(os.path.abspath(__file__))
class HumanContext:
def __init__(self):
self._fps = 50 # 20 ms per frame
self._image_size = 128
self._image_size = 288
self._batch_size = 16
self._sample_rate = 16000
self._stride_left_size = 10