Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import torch
|
| 2 |
+
import gradio as gr
|
| 3 |
+
from model import ECAPA_gender
|
| 4 |
+
|
| 5 |
+
model = ECAPA_gender.from_pretrained("Beijuka/voice-gender-classifier")
|
| 6 |
+
model.eval()
|
| 7 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
| 8 |
+
model.to(device)
|
| 9 |
+
|
| 10 |
+
def predict_gender_confidence(audio_file):
|
| 11 |
+
if audio_file is None:
|
| 12 |
+
return "No audio provided"
|
| 13 |
+
|
| 14 |
+
try:
|
| 15 |
+
# Load audio
|
| 16 |
+
audio = model.load_audio(audio_file.name if hasattr(audio_file, "name") else audio_file)
|
| 17 |
+
audio = audio.to(device)
|
| 18 |
+
|
| 19 |
+
# Forward pass
|
| 20 |
+
with torch.no_grad():
|
| 21 |
+
logits = model.forward(audio)
|
| 22 |
+
probs = torch.softmax(logits, dim=1).cpu().numpy()[0]
|
| 23 |
+
pred_idx = logits.argmax(dim=1).item()
|
| 24 |
+
gender_pred = model.pred2gender[pred_idx].capitalize()
|
| 25 |
+
confidence = probs[pred_idx] * 100
|
| 26 |
+
|
| 27 |
+
return f"{gender_pred} — {confidence:.1f}% confidence"
|
| 28 |
+
|
| 29 |
+
except Exception as e:
|
| 30 |
+
return f"Error: {e}"
|
| 31 |
+
|
| 32 |
+
iface = gr.Interface(
|
| 33 |
+
fn=predict_gender_confidence,
|
| 34 |
+
inputs=gr.Audio(type="filepath", label="Upload audio file", sources=["upload"]),
|
| 35 |
+
outputs=gr.Textbox(label="Predicted Gender with Confidence"),
|
| 36 |
+
title="Voice Gender Classifier",
|
| 37 |
+
description="Upload an audio file and the model predicts speaker gender with confidence.",
|
| 38 |
+
allow_flagging="never"
|
| 39 |
+
)
|
| 40 |
+
|
| 41 |
+
iface.launch(share=True)
|