Spaces:
Sleeping
Sleeping
Commit
·
23ba234
1
Parent(s):
0073001
app.py
CHANGED
|
@@ -6,11 +6,19 @@ import numpy as np
|
|
| 6 |
import tempfile
|
| 7 |
import os
|
| 8 |
import warnings
|
|
|
|
| 9 |
|
| 10 |
warnings.filterwarnings("ignore")
|
| 11 |
|
| 12 |
app = FastAPI()
|
| 13 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 14 |
def extract_audio_features(audio_file_path):
|
| 15 |
# Load the audio file using soundfile
|
| 16 |
waveform, sample_rate = sf.read(audio_file_path)
|
|
@@ -24,7 +32,6 @@ def extract_audio_features(audio_file_path):
|
|
| 24 |
mfccs = np.mean(np.abs(np.fft.fft(waveform)[:13]), axis=0) # Simplified MFCC-like features
|
| 25 |
|
| 26 |
# Placeholder for speech rate and fundamental frequency
|
| 27 |
-
# Speech rate and pitch extraction would require more complex DSP techniques or external libraries.
|
| 28 |
speech_rate = 4.0 # Arbitrary placeholder value for speech rate
|
| 29 |
f0 = np.mean(np.abs(np.diff(waveform))) * sample_rate / (2 * np.pi) # Rough pitch estimate
|
| 30 |
|
|
@@ -77,27 +84,31 @@ async def analyze_stress(
|
|
| 77 |
# Handle audio file analysis
|
| 78 |
if file or file_path:
|
| 79 |
if file:
|
| 80 |
-
if not file.filename.endswith(".wav"):
|
| 81 |
-
raise HTTPException(status_code=400, detail="Only .wav files are supported.")
|
| 82 |
-
with tempfile.NamedTemporaryFile(delete=False, suffix=
|
| 83 |
temp_file.write(await file.read())
|
| 84 |
-
|
| 85 |
else:
|
| 86 |
-
if not file_path.endswith(".wav"):
|
| 87 |
-
raise HTTPException(status_code=400, detail="Only .wav files are supported.")
|
| 88 |
if not os.path.exists(file_path):
|
| 89 |
raise HTTPException(status_code=400, detail="File path does not exist.")
|
| 90 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 91 |
|
| 92 |
try:
|
| 93 |
-
result = analyze_voice_stress(
|
| 94 |
return JSONResponse(content=result)
|
| 95 |
except Exception as e:
|
| 96 |
raise HTTPException(status_code=500, detail=str(e))
|
| 97 |
finally:
|
| 98 |
# Clean up temporary files
|
| 99 |
if file:
|
| 100 |
-
os.remove(
|
| 101 |
|
| 102 |
# Handle text analysis
|
| 103 |
elif text:
|
|
|
|
| 6 |
import tempfile
|
| 7 |
import os
|
| 8 |
import warnings
|
| 9 |
+
from pydub import AudioSegment
|
| 10 |
|
| 11 |
warnings.filterwarnings("ignore")
|
| 12 |
|
| 13 |
app = FastAPI()
|
| 14 |
|
| 15 |
+
def convert_mp3_to_wav(mp3_path):
|
| 16 |
+
# Convert MP3 to WAV
|
| 17 |
+
sound = AudioSegment.from_mp3(mp3_path)
|
| 18 |
+
wav_path = mp3_path.replace(".mp3", ".wav")
|
| 19 |
+
sound.export(wav_path, format="wav")
|
| 20 |
+
return wav_path
|
| 21 |
+
|
| 22 |
def extract_audio_features(audio_file_path):
|
| 23 |
# Load the audio file using soundfile
|
| 24 |
waveform, sample_rate = sf.read(audio_file_path)
|
|
|
|
| 32 |
mfccs = np.mean(np.abs(np.fft.fft(waveform)[:13]), axis=0) # Simplified MFCC-like features
|
| 33 |
|
| 34 |
# Placeholder for speech rate and fundamental frequency
|
|
|
|
| 35 |
speech_rate = 4.0 # Arbitrary placeholder value for speech rate
|
| 36 |
f0 = np.mean(np.abs(np.diff(waveform))) * sample_rate / (2 * np.pi) # Rough pitch estimate
|
| 37 |
|
|
|
|
| 84 |
# Handle audio file analysis
|
| 85 |
if file or file_path:
|
| 86 |
if file:
|
| 87 |
+
if not (file.filename.endswith(".wav") or file.filename.endswith(".mp3")):
|
| 88 |
+
raise HTTPException(status_code=400, detail="Only .wav and .mp3 files are supported.")
|
| 89 |
+
with tempfile.NamedTemporaryFile(delete=False, suffix=os.path.splitext(file.filename)[-1]) as temp_file:
|
| 90 |
temp_file.write(await file.read())
|
| 91 |
+
temp_audio_path = temp_file.name
|
| 92 |
else:
|
| 93 |
+
if not (file_path.endswith(".wav") or file_path.endswith(".mp3")):
|
| 94 |
+
raise HTTPException(status_code=400, detail="Only .wav and .mp3 files are supported.")
|
| 95 |
if not os.path.exists(file_path):
|
| 96 |
raise HTTPException(status_code=400, detail="File path does not exist.")
|
| 97 |
+
temp_audio_path = file_path
|
| 98 |
+
|
| 99 |
+
# Convert MP3 to WAV if needed
|
| 100 |
+
if temp_audio_path.endswith(".mp3"):
|
| 101 |
+
temp_audio_path = convert_mp3_to_wav(temp_audio_path)
|
| 102 |
|
| 103 |
try:
|
| 104 |
+
result = analyze_voice_stress(temp_audio_path)
|
| 105 |
return JSONResponse(content=result)
|
| 106 |
except Exception as e:
|
| 107 |
raise HTTPException(status_code=500, detail=str(e))
|
| 108 |
finally:
|
| 109 |
# Clean up temporary files
|
| 110 |
if file:
|
| 111 |
+
os.remove(temp_audio_path)
|
| 112 |
|
| 113 |
# Handle text analysis
|
| 114 |
elif text:
|