Update app.py
Browse files
app.py
CHANGED
@@ -4,18 +4,18 @@ import os
|
|
4 |
token = os.environ.get('HF_TOKEN')
|
5 |
whisper_to_gpt = gr.Blocks.load(name="spaces/fffiloni/whisper-to-chatGPT")
|
6 |
tts = gr.Interface.load(name="spaces/StevenLimcorn/fastspeech2-TTS")
|
7 |
-
|
8 |
|
9 |
def infer(audio):
|
10 |
gpt_response = whisper_to_gpt(audio, "translate", fn_index=0)
|
11 |
#print(gpt_response)
|
12 |
audio_response = tts(gpt_response[1], "Fastspeech2 + Melgan", fn_index=0)
|
13 |
#image = Image.open(r"wise_woman_portrait.png")
|
14 |
-
|
15 |
-
return audio_response
|
16 |
|
17 |
inputs = gr.Audio(source="microphone",type="filepath")
|
18 |
-
outputs = [gr.Audio()]
|
19 |
|
20 |
demo = gr.Interface(fn=infer, inputs=inputs, outputs=outputs)
|
21 |
demo.launch()
|
|
|
4 |
token = os.environ.get('HF_TOKEN')
|
5 |
whisper_to_gpt = gr.Blocks.load(name="spaces/fffiloni/whisper-to-chatGPT")
|
6 |
tts = gr.Interface.load(name="spaces/StevenLimcorn/fastspeech2-TTS")
|
7 |
+
talking_face = gr.Blocks.load(name="spaces/fffiloni/one-shot-talking-face", api_key=token)
|
8 |
|
9 |
def infer(audio):
|
10 |
gpt_response = whisper_to_gpt(audio, "translate", fn_index=0)
|
11 |
#print(gpt_response)
|
12 |
audio_response = tts(gpt_response[1], "Fastspeech2 + Melgan", fn_index=0)
|
13 |
#image = Image.open(r"wise_woman_portrait.png")
|
14 |
+
portrait = talking_face("wise_woman_portrait.png", audio_response, fn_index=0)
|
15 |
+
return audio_response, portrait
|
16 |
|
17 |
inputs = gr.Audio(source="microphone",type="filepath")
|
18 |
+
outputs = [gr.Audio(), gr.Video()]
|
19 |
|
20 |
demo = gr.Interface(fn=infer, inputs=inputs, outputs=outputs)
|
21 |
demo.launch()
|