File size: 17,194 Bytes
28fc64c e1e69e9 28fc64c 89bed5d e038003 89bed5d e1e69e9 f84d4de e800564 28fc64c e1e69e9 28fc64c f84d4de e800564 e1e69e9 e800564 28fc64c e1e69e9 28fc64c e1e69e9 e800564 e1e69e9 f900f76 e1e69e9 f900f76 e1e69e9 f900f76 e1e69e9 f900f76 e1e69e9 f900f76 e1e69e9 f900f76 e1e69e9 f900f76 e1e69e9 f900f76 e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e800564 28fc64c e800564 28fc64c f84d4de e800564 28fc64c f84d4de f900f76 e1e69e9 28fc64c e800564 28fc64c e1e69e9 28fc64c e800564 28fc64c e800564 28fc64c e800564 28fc64c e800564 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e800564 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e800564 28fc64c e1e69e9 28fc64c e800564 e1e69e9 28fc64c f900f76 28fc64c e1e69e9 28fc64c f84d4de 28fc64c e1e69e9 e800564 28fc64c e1e69e9 28fc64c f84d4de 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 e800564 28fc64c e1e69e9 bf5780d e1e69e9 28fc64c bf5780d 28fc64c bf5780d e1e69e9 28fc64c e1e69e9 bf5780d 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c bf5780d e1e69e9 bf5780d e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 bf5780d e1e69e9 bf5780d 28fc64c e1e69e9 bf5780d e1e69e9 bf5780d 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c e1e69e9 28fc64c bf5780d 28fc64c e1e69e9 3f2b9ca 28fc64c 3f2b9ca e1e69e9 3f2b9ca e1e69e9 | 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 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 | #!/usr/bin/env python3
"""
Lightweight Aphasia Classification App
Optimized for Hugging Face Spaces with lazy loading and fallbacks
"""
import os
# Configure environment for CPU-only and memory optimization
os.environ['PYTORCH_CUDA_ALLOC_CONF'] = 'max_split_size_mb:512'
os.environ['OMP_NUM_THREADS'] = '2' # Limit CPU threads
os.environ['MKL_NUM_THREADS'] = '2'
os.environ['NUMEXPR_NUM_THREADS'] = '2'
os.environ['TOKENIZERS_PARALLELISM'] = 'false' # Avoid tokenizer warnings
# Batchalign specific settings
os.environ['BATCHALIGN_CACHE'] = '/tmp/batchalign_cache'
os.environ['HF_HUB_CACHE'] = '/tmp/hf_cache' # Use tmp for model cache
os.environ['TRANSFORMERS_CACHE'] = '/tmp/transformers_cache'
# Whisper settings for CPU optimization
os.environ['WHISPER_CACHE'] = '/tmp/whisper_cache'
print("🔧 Environment configured for CPU-only processing")
print("💾 Model caches set to /tmp/ to save space")
from flask import Flask, request, render_template_string, jsonify
import os
import tempfile
import logging
import json
import threading
import time
from pathlib import Path
# Set up logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)
app = Flask(__name__)
app.config['MAX_CONTENT_LENGTH'] = 50 * 1024 * 1024 # 50MB max (reduced)
print("🚀 Starting Lightweight Aphasia Classification System")
# Global state
MODULES = {}
MODELS_LOADED = False
LOADING_STATUS = "Starting up..."
def lazy_import_modules():
"""Import modules only when needed"""
global MODULES, MODELS_LOADED, LOADING_STATUS
if MODELS_LOADED:
return True
try:
LOADING_STATUS = "Loading audio processing..."
logger.info("Importing utils_audio...")
from utils_audio import convert_to_wav
MODULES['convert_to_wav'] = convert_to_wav
logger.info("✓ Audio processing loaded")
LOADING_STATUS = "Loading speech analysis..."
logger.info("Importing to_cha...")
from to_cha import to_cha_from_wav
MODULES['to_cha_from_wav'] = to_cha_from_wav
logger.info("✓ Speech analysis loaded")
LOADING_STATUS = "Loading data conversion..."
logger.info("Importing cha_json...")
from cha_json import cha_to_json_file
MODULES['cha_to_json_file'] = cha_to_json_file
logger.info("✓ Data conversion loaded")
LOADING_STATUS = "Loading AI model..."
logger.info("Importing output...")
from output import predict_from_chajson
MODULES['predict_from_chajson'] = predict_from_chajson
logger.info("✓ AI model loaded")
MODELS_LOADED = True
LOADING_STATUS = "Ready!"
logger.info("🎉 All modules loaded successfully!")
return True
except Exception as e:
logger.error(f"Failed to load modules: {e}")
LOADING_STATUS = f"Error: {str(e)}"
return False
def background_loader():
"""Load modules in background thread"""
logger.info("Starting background module loading...")
lazy_import_modules()
# Start loading modules in background
loading_thread = threading.Thread(target=background_loader, daemon=True)
loading_thread.start()
# HTML Template (simplified)
HTML_TEMPLATE = """
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>🧠 Aphasia Classification</title>
<style>
body {
font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
min-height: 100vh;
padding: 20px;
margin: 0;
}
.container {
max-width: 800px;
margin: 0 auto;
background: white;
border-radius: 20px;
box-shadow: 0 20px 60px rgba(0,0,0,0.1);
overflow: hidden;
}
.header {
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
color: white;
padding: 40px 30px;
text-align: center;
}
.content {
padding: 40px 30px;
}
.status {
background: #f8f9fa;
border-radius: 10px;
padding: 20px;
margin-bottom: 30px;
border-left: 4px solid #28a745;
}
.status.loading {
border-left-color: #ffc107;
}
.status.error {
border-left-color: #dc3545;
}
.upload-section {
background: #f8f9fa;
border-radius: 15px;
padding: 30px;
text-align: center;
margin-bottom: 30px;
}
.file-input {
display: none;
}
.file-label {
display: inline-block;
background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
color: white;
padding: 15px 30px;
border-radius: 50px;
cursor: pointer;
font-weight: 600;
transition: transform 0.2s ease;
}
.file-label:hover {
transform: translateY(-2px);
}
.analyze-btn {
background: #28a745;
color: white;
border: none;
padding: 15px 40px;
border-radius: 50px;
font-weight: 600;
cursor: pointer;
margin-top: 20px;
transition: all 0.2s ease;
}
.analyze-btn:disabled {
background: #6c757d;
cursor: not-allowed;
}
.results {
background: #f8f9fa;
border-radius: 15px;
padding: 30px;
margin-top: 30px;
display: none;
white-space: pre-wrap;
font-family: monospace;
}
.loading {
text-align: center;
padding: 40px;
display: none;
}
.spinner {
border: 4px solid #f3f3f3;
border-top: 4px solid #667eea;
border-radius: 50%;
width: 50px;
height: 50px;
animation: spin 1s linear infinite;
margin: 0 auto 20px;
}
@keyframes spin {
0% { transform: rotate(0deg); }
100% { transform: rotate(360deg); }
}
.refresh-btn {
background: #17a2b8;
color: white;
border: none;
padding: 10px 20px;
border-radius: 25px;
cursor: pointer;
margin-left: 10px;
}
</style>
</head>
<body>
<div class="container">
<div class="header">
<h1>🧠 Aphasia Classification</h1>
<p>AI-powered speech analysis for aphasia identification</p>
</div>
<div class="content">
<div class="status" id="statusBox">
<h3 id="statusTitle">🔄 System Status</h3>
<div id="statusText">{{ status_message }}</div>
<button class="refresh-btn" onclick="checkStatus()">Refresh Status</button>
</div>
<div class="upload-section">
<h3>📁 Upload Audio File</h3>
<p>Upload speech audio for aphasia classification</p>
<form id="uploadForm" enctype="multipart/form-data">
<input type="file" id="audioFile" name="audio" class="file-input" accept="audio/*" required>
<label for="audioFile" class="file-label">
🎵 Choose Audio File
</label>
<br>
<button type="submit" class="analyze-btn" id="analyzeBtn">
🔍 Analyze Speech
</button>
</form>
<p style="color: #666; margin-top: 15px; font-size: 0.9em;">
Supported: MP3, WAV, M4A (max 50MB)
</p>
</div>
<div class="loading" id="loading">
<div class="spinner"></div>
<h3>🔄 Processing Audio...</h3>
<p>This may take 2-5 minutes. Please be patient.</p>
</div>
<div class="results" id="results"></div>
</div>
</div>
<script>
// Check status periodically
function checkStatus() {
fetch('/status')
.then(response => response.json())
.then(data => {
const statusBox = document.getElementById('statusBox');
const statusTitle = document.getElementById('statusTitle');
const statusText = document.getElementById('statusText');
if (data.ready) {
statusBox.className = 'status';
statusTitle.textContent = '🟢 System Ready';
statusText.textContent = 'All components loaded. Ready to process audio files.';
} else {
statusBox.className = 'status loading';
statusTitle.textContent = '🟡 Loading...';
statusText.textContent = data.status;
}
})
.catch(error => {
const statusBox = document.getElementById('statusBox');
statusBox.className = 'status error';
document.getElementById('statusTitle').textContent = '🔴 Error';
document.getElementById('statusText').textContent = 'Failed to check status';
});
}
// Check status every 5 seconds
setInterval(checkStatus, 5000);
// Form submission
document.getElementById('uploadForm').addEventListener('submit', async function(e) {
e.preventDefault();
const fileInput = document.getElementById('audioFile');
const loading = document.getElementById('loading');
const results = document.getElementById('results');
const analyzeBtn = document.getElementById('analyzeBtn');
if (!fileInput.files[0]) {
alert('Please select an audio file');
return;
}
// Check if system is ready
const statusCheck = await fetch('/status');
const status = await statusCheck.json();
if (!status.ready) {
alert('System is still loading. Please wait and try again.');
return;
}
// Show loading
loading.style.display = 'block';
results.style.display = 'none';
analyzeBtn.disabled = true;
analyzeBtn.textContent = 'Processing...';
try {
const formData = new FormData();
formData.append('audio', fileInput.files[0]);
const response = await fetch('/analyze', {
method: 'POST',
body: formData
});
const data = await response.json();
loading.style.display = 'none';
if (data.success) {
results.textContent = data.result;
results.style.borderLeft = '4px solid #28a745';
} else {
results.textContent = 'Error: ' + data.error;
results.style.borderLeft = '4px solid #dc3545';
}
results.style.display = 'block';
} catch (error) {
loading.style.display = 'none';
results.textContent = 'Network error: ' + error.message;
results.style.borderLeft = '4px solid #dc3545';
results.style.display = 'block';
}
analyzeBtn.disabled = false;
analyzeBtn.textContent = '🔍 Analyze Speech';
});
// File selection feedback
document.getElementById('audioFile').addEventListener('change', function(e) {
const label = document.querySelector('.file-label');
if (e.target.files[0]) {
label.textContent = '✓ ' + e.target.files[0].name;
} else {
label.textContent = '🎵 Choose Audio File';
}
});
</script>
</body>
</html>
"""
@app.route('/')
def index():
"""Main page"""
return render_template_string(HTML_TEMPLATE, status_message=LOADING_STATUS)
@app.route('/status')
def status():
"""Status check endpoint"""
return jsonify({
'ready': MODELS_LOADED,
'status': LOADING_STATUS,
'modules_loaded': len(MODULES)
})
@app.route('/analyze', methods=['POST'])
def analyze_audio():
"""Process uploaded audio - only if models are loaded"""
try:
# Check if system is ready
if not MODELS_LOADED:
return jsonify({
'success': False,
'error': f'System still loading: {LOADING_STATUS}'
})
# Check file upload
if 'audio' not in request.files:
return jsonify({'success': False, 'error': 'No audio file uploaded'})
audio_file = request.files['audio']
if audio_file.filename == '':
return jsonify({'success': False, 'error': 'No file selected'})
# Save uploaded file
with tempfile.NamedTemporaryFile(delete=False, suffix=os.path.splitext(audio_file.filename)[1]) as tmp_file:
audio_file.save(tmp_file.name)
temp_path = tmp_file.name
try:
logger.info("🎵 Starting audio processing...")
# Step 1: Convert to WAV
logger.info("Converting to WAV...")
wav_path = MODULES['convert_to_wav'](temp_path, sr=16000, mono=True)
# Step 2: Generate CHA
logger.info("Generating CHA file...")
cha_path = MODULES['to_cha_from_wav'](wav_path, lang="eng")
# Step 3: Convert to JSON
logger.info("Converting to JSON...")
json_path, _ = MODULES['cha_to_json_file'](cha_path)
# Step 4: Classification
logger.info("Running classification...")
results = MODULES['predict_from_chajson'](".", json_path, output_file=None)
# Cleanup
for temp_file in [temp_path, wav_path, cha_path, json_path]:
try:
os.unlink(temp_file)
except:
pass
# Format results
if "predictions" in results and results["predictions"]:
pred = results["predictions"][0]
classification = pred["prediction"]["predicted_class"]
confidence = pred["prediction"]["confidence_percentage"]
description = pred["class_description"]["name"]
severity = pred["additional_predictions"]["predicted_severity_level"]
fluency = pred["additional_predictions"]["fluency_rating"]
result_text = f"""🧠 APHASIA CLASSIFICATION RESULTS
🎯 Classification: {classification}
📊 Confidence: {confidence}
📋 Type: {description}
📈 Severity: {severity}/3
🗣️ Fluency: {fluency}
📊 Top 3 Probabilities:"""
prob_dist = pred["probability_distribution"]
for i, (atype, info) in enumerate(list(prob_dist.items())[:3], 1):
result_text += f"\n{i}. {atype}: {info['percentage']}"
result_text += f"""
📝 Description:
{pred["class_description"]["description"]}
✅ Processing completed successfully!
"""
return jsonify({'success': True, 'result': result_text})
else:
return jsonify({'success': False, 'error': 'No predictions generated'})
except Exception as e:
# Cleanup on error
try:
os.unlink(temp_path)
except:
pass
raise e
except Exception as e:
logger.error(f"Processing error: {e}")
return jsonify({'success': False, 'error': str(e)})
if __name__ == '__main__':
port = int(os.environ.get('PORT', 7860))
print(f"🚀 Starting on port {port}")
print("🔄 Models loading in background...")
app.run(host='0.0.0.0', port=port, debug=False, threaded=True) |