Spaces:
Running
Running
File size: 2,117 Bytes
df2accb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 |
import gradio as gr
SUPPORTED_TARGET_SINGERS = {
"Adele": "vocalist_l1_Adele",
"Beyonce": "vocalist_l1_Beyonce",
"Bruno Mars": "vocalist_l1_BrunoMars",
"John Mayer": "vocalist_l1_JohnMayer",
"Michael Jackson": "vocalist_l1_MichaelJackson",
"Taylor Swift": "vocalist_l1_TaylorSwift",
"Jacky Cheung 张学友": "vocalist_l1_张学友",
"Jian Li 李健": "vocalist_l1_李健",
"Feng Wang 汪峰": "vocalist_l1_汪峰",
"Faye Wong 王菲": "vocalist_l1_王菲",
"Yijie Shi 石倚洁": "vocalist_l1_石倚洁",
"Tsai Chin 蔡琴": "vocalist_l1_蔡琴",
"Ying Na 那英": "vocalist_l1_那英",
"Eason Chan 陈奕迅": "vocalist_l1_陈奕迅",
"David Tao 陶喆": "vocalist_l1_陶喆",
}
def svc_inference(
source_audio,
target_singer,
diffusion_steps=1000,
key_shift_mode="auto",
key_shift_num=0,
):
pass
demo_inputs = [
gr.Audio(
sources=["upload", "microphone"],
label="Upload (or record) a song you want to listen",
),
gr.Radio(
choices=list(SUPPORTED_TARGET_SINGERS.keys()),
label="Target Singer",
value="Jian Li 李健",
),
gr.Slider(
1,
1000,
value=1000,
step=1,
label="Diffusion Inference Steps",
info="As the step number increases, the synthesis quality will be better while the inference speed will be lower",
),
gr.Radio(
choices=["Auto Shift", "Key Shift"],
value="Auto Shift",
label="Pitch Shift Control",
info='If you want to control the specific pitch shift value, you need to choose "Key Shift"',
),
gr.Slider(
-6,
6,
value=0,
step=1,
label="Key Shift Values",
info='How many semitones you want to transpose. This parameter will work only if you choose "Key Shift"',
),
]
demo_outputs = gr.Audio(label="")
demo = gr.Interface(
fn=svc_inference,
inputs=demo_inputs,
outputs=demo_outputs,
title="Amphion Singing Voice Conversion",
)
if __name__ == "__main__":
demo.launch(show_api=False)
|