Spaces:
Paused
Paused
File size: 3,398 Bytes
1e3bdcd ed64ea8 1e3bdcd | 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 | import os
import argparse
import pandas as pd
from tqdm import tqdm
import warnings
# Suppress warnings
warnings.filterwarnings("ignore")
from metrics.srmr_metric import calculate_srmr
from metrics.sigmos_metric import calculate_sigmos
from metrics.vqscore_metric import calculate_vqscore
from metrics.wvmos_metric import calculate_wvmos
METRIC_DESCRIPTIONS = {
'SRMR': 'Technical measurement of reverberation and room acoustics',
'SIGMOS_DISC': 'Audio continuity and smoothness',
'VQScore': 'Overall voice quality assessment',
'WVMOS': 'Predicted subjective quality rating',
'SIGMOS_OVRL': 'Comprehensive overall audio quality',
'SIGMOS_REVERB': 'Perceived reverberation quality'
}
import config
THRESHOLDS = config.THRESHOLDS
def evaluate_file(file_path):
# Calculate all scores first
scores = {}
# SRMR
scores['SRMR'] = calculate_srmr(file_path)
# SIGMOS
sigmos_scores = calculate_sigmos(file_path)
if sigmos_scores:
scores.update(sigmos_scores)
else:
scores['SIGMOS_DISC'] = None
scores['SIGMOS_OVRL'] = None
scores['SIGMOS_REVERB'] = None
# VQScore
scores['VQScore'] = calculate_vqscore(file_path)
# WVMOS
scores['WVMOS'] = calculate_wvmos(file_path)
# Build list of rows for this file
rows = []
filename = os.path.basename(file_path)
for metric, threshold in THRESHOLDS.items():
score = scores.get(metric)
if score is None:
status = 'ERROR'
elif score >= threshold:
status = 'PASS'
else:
status = 'FAIL'
row = {
'Filename': filename,
'Metric': metric,
'Description': METRIC_DESCRIPTIONS.get(metric, ''),
'Threshold': threshold,
'Score': score,
'PASS OR FAIL': status
}
rows.append(row)
return rows
def main():
parser = argparse.ArgumentParser(description='TTS Audio Quality Evaluation')
parser.add_argument('input_path', help='Path to audio file or directory')
parser.add_argument('--output', default='evaluation_report.csv', help='Output CSV file')
args = parser.parse_args()
input_path = args.input_path
files = []
if os.path.isdir(input_path):
for root, dirs, filenames in os.walk(input_path):
for filename in filenames:
if filename.lower().endswith(('.wav', '.mp3', '.flac')):
files.append(os.path.join(root, filename))
elif os.path.isfile(input_path):
files.append(input_path)
else:
print(f"Invalid path: {input_path}")
return
print(f"Found {len(files)} files to evaluate.")
all_rows = []
for file_path in tqdm(files):
try:
file_rows = evaluate_file(file_path)
all_rows.extend(file_rows)
except Exception as e:
import traceback
traceback.print_exc()
print(f"Failed to process {file_path}: {e}")
df = pd.DataFrame(all_rows)
# Ensure column order
cols = ['Filename', 'Metric', 'Description', 'Threshold', 'Score', 'PASS OR FAIL']
df = df[cols]
df.to_csv(args.output, index=False)
print(f"Report saved to {args.output}")
if __name__ == '__main__':
main()
|