Update app.py
Browse files
app.py
CHANGED
|
@@ -1,4 +1,4 @@
|
|
| 1 |
-
from fastapi import FastAPI, HTTPException
|
| 2 |
from fastapi.responses import JSONResponse, FileResponse
|
| 3 |
from pydantic import BaseModel
|
| 4 |
import numpy as np
|
|
@@ -10,12 +10,13 @@ import torch
|
|
| 10 |
import librosa
|
| 11 |
from transformers import Wav2Vec2ForCTC, AutoProcessor
|
| 12 |
from pathlib import Path
|
|
|
|
|
|
|
| 13 |
|
| 14 |
# Import functions from other modules
|
| 15 |
from asr import transcribe, ASR_LANGUAGES
|
| 16 |
from tts import synthesize, TTS_LANGUAGES
|
| 17 |
from lid import identify
|
| 18 |
-
|
| 19 |
from asr import ASR_SAMPLING_RATE, transcribe
|
| 20 |
|
| 21 |
# Configure logging
|
|
@@ -26,7 +27,7 @@ app = FastAPI(title="MMS: Scaling Speech Technology to 1000+ languages")
|
|
| 26 |
|
| 27 |
# Define request models
|
| 28 |
class AudioRequest(BaseModel):
|
| 29 |
-
audio: str # Base64 encoded audio data
|
| 30 |
language: str
|
| 31 |
|
| 32 |
class TTSRequest(BaseModel):
|
|
@@ -34,19 +35,42 @@ class TTSRequest(BaseModel):
|
|
| 34 |
language: str
|
| 35 |
speed: float
|
| 36 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
@app.post("/transcribe")
|
| 38 |
async def transcribe_audio(request: AudioRequest):
|
| 39 |
try:
|
| 40 |
-
|
| 41 |
-
audio_array, sample_rate =
|
|
|
|
| 42 |
# Convert to mono if stereo
|
| 43 |
if len(audio_array.shape) > 1:
|
| 44 |
audio_array = audio_array.mean(axis=1)
|
|
|
|
| 45 |
# Ensure audio_array is float32
|
| 46 |
audio_array = audio_array.astype(np.float32)
|
|
|
|
| 47 |
# Resample if necessary
|
| 48 |
if sample_rate != ASR_SAMPLING_RATE:
|
| 49 |
audio_array = librosa.resample(audio_array, orig_sr=sample_rate, target_sr=ASR_SAMPLING_RATE)
|
|
|
|
| 50 |
result = transcribe(audio_array, request.language)
|
| 51 |
return JSONResponse(content={"transcription": result})
|
| 52 |
except Exception as e:
|
|
@@ -57,15 +81,13 @@ async def transcribe_audio(request: AudioRequest):
|
|
| 57 |
async def synthesize_speech(request: TTSRequest):
|
| 58 |
try:
|
| 59 |
audio, filtered_text = synthesize(request.text, request.language, request.speed)
|
| 60 |
-
|
| 61 |
# Convert numpy array to bytes
|
| 62 |
buffer = io.BytesIO()
|
| 63 |
sf.write(buffer, audio, 22050, format='wav')
|
| 64 |
buffer.seek(0)
|
| 65 |
-
|
| 66 |
return FileResponse(
|
| 67 |
-
buffer,
|
| 68 |
-
media_type="audio/wav",
|
| 69 |
headers={"Content-Disposition": "attachment; filename=synthesized_audio.wav"}
|
| 70 |
)
|
| 71 |
except Exception as e:
|
|
@@ -75,9 +97,8 @@ async def synthesize_speech(request: TTSRequest):
|
|
| 75 |
@app.post("/identify")
|
| 76 |
async def identify_language(request: AudioRequest):
|
| 77 |
try:
|
| 78 |
-
|
| 79 |
-
audio_array, sample_rate =
|
| 80 |
-
|
| 81 |
result = identify(audio_array)
|
| 82 |
return JSONResponse(content={"language_identification": result})
|
| 83 |
except Exception as e:
|
|
|
|
| 1 |
+
from fastapi import FastAPI, HTTPException, UploadFile, File
|
| 2 |
from fastapi.responses import JSONResponse, FileResponse
|
| 3 |
from pydantic import BaseModel
|
| 4 |
import numpy as np
|
|
|
|
| 10 |
import librosa
|
| 11 |
from transformers import Wav2Vec2ForCTC, AutoProcessor
|
| 12 |
from pathlib import Path
|
| 13 |
+
from moviepy.editor import VideoFileClip
|
| 14 |
+
import magic # For MIME type detection
|
| 15 |
|
| 16 |
# Import functions from other modules
|
| 17 |
from asr import transcribe, ASR_LANGUAGES
|
| 18 |
from tts import synthesize, TTS_LANGUAGES
|
| 19 |
from lid import identify
|
|
|
|
| 20 |
from asr import ASR_SAMPLING_RATE, transcribe
|
| 21 |
|
| 22 |
# Configure logging
|
|
|
|
| 27 |
|
| 28 |
# Define request models
|
| 29 |
class AudioRequest(BaseModel):
|
| 30 |
+
audio: str # Base64 encoded audio or video data
|
| 31 |
language: str
|
| 32 |
|
| 33 |
class TTSRequest(BaseModel):
|
|
|
|
| 35 |
language: str
|
| 36 |
speed: float
|
| 37 |
|
| 38 |
+
def detect_mime_type(input_bytes):
|
| 39 |
+
mime = magic.Magic(mime=True)
|
| 40 |
+
return mime.from_buffer(input_bytes)
|
| 41 |
+
|
| 42 |
+
def extract_audio(input_bytes):
|
| 43 |
+
mime_type = detect_mime_type(input_bytes)
|
| 44 |
+
|
| 45 |
+
if mime_type.startswith('audio/'):
|
| 46 |
+
return sf.read(io.BytesIO(input_bytes))
|
| 47 |
+
elif mime_type.startswith('video/'):
|
| 48 |
+
with io.BytesIO(input_bytes) as f:
|
| 49 |
+
video = VideoFileClip(f.name)
|
| 50 |
+
audio = video.audio
|
| 51 |
+
audio_array = audio.to_soundarray()
|
| 52 |
+
sample_rate = audio.fps
|
| 53 |
+
return audio_array, sample_rate
|
| 54 |
+
else:
|
| 55 |
+
raise ValueError(f"Unsupported MIME type: {mime_type}")
|
| 56 |
+
|
| 57 |
@app.post("/transcribe")
|
| 58 |
async def transcribe_audio(request: AudioRequest):
|
| 59 |
try:
|
| 60 |
+
input_bytes = base64.b64decode(request.audio)
|
| 61 |
+
audio_array, sample_rate = extract_audio(input_bytes)
|
| 62 |
+
|
| 63 |
# Convert to mono if stereo
|
| 64 |
if len(audio_array.shape) > 1:
|
| 65 |
audio_array = audio_array.mean(axis=1)
|
| 66 |
+
|
| 67 |
# Ensure audio_array is float32
|
| 68 |
audio_array = audio_array.astype(np.float32)
|
| 69 |
+
|
| 70 |
# Resample if necessary
|
| 71 |
if sample_rate != ASR_SAMPLING_RATE:
|
| 72 |
audio_array = librosa.resample(audio_array, orig_sr=sample_rate, target_sr=ASR_SAMPLING_RATE)
|
| 73 |
+
|
| 74 |
result = transcribe(audio_array, request.language)
|
| 75 |
return JSONResponse(content={"transcription": result})
|
| 76 |
except Exception as e:
|
|
|
|
| 81 |
async def synthesize_speech(request: TTSRequest):
|
| 82 |
try:
|
| 83 |
audio, filtered_text = synthesize(request.text, request.language, request.speed)
|
|
|
|
| 84 |
# Convert numpy array to bytes
|
| 85 |
buffer = io.BytesIO()
|
| 86 |
sf.write(buffer, audio, 22050, format='wav')
|
| 87 |
buffer.seek(0)
|
|
|
|
| 88 |
return FileResponse(
|
| 89 |
+
buffer,
|
| 90 |
+
media_type="audio/wav",
|
| 91 |
headers={"Content-Disposition": "attachment; filename=synthesized_audio.wav"}
|
| 92 |
)
|
| 93 |
except Exception as e:
|
|
|
|
| 97 |
@app.post("/identify")
|
| 98 |
async def identify_language(request: AudioRequest):
|
| 99 |
try:
|
| 100 |
+
input_bytes = base64.b64decode(request.audio)
|
| 101 |
+
audio_array, sample_rate = extract_audio(input_bytes)
|
|
|
|
| 102 |
result = identify(audio_array)
|
| 103 |
return JSONResponse(content={"language_identification": result})
|
| 104 |
except Exception as e:
|