| import pandas as pd |
| import os |
| import torch |
| import torchaudio |
| import soundfile as sf |
| from tqdm import tqdm |
|
|
| target_sample_rate = 16000 |
|
|
| df = pd.read_parquet('data.parquet') |
|
|
| |
| df = df.rename(columns={ |
| 'query_audio': 'query_audio_path', |
| 'doc_audio': 'document_audio_path', |
| 'query_duration': 'query_audio_duration', |
| 'doc_duration': 'document_audio_duration', |
| }) |
|
|
| |
| base_dir = os.path.dirname(os.path.abspath(__file__)) if '__file__' in globals() else os.getcwd() |
|
|
| def check_and_resample_audio(audio_path, base_dir): |
| """ |
| 检查音频文件的采样率,如果不是16kHz则重采样 |
| |
| Args: |
| audio_path: 音频文件的相对路径(在parquet中存储的路径) |
| base_dir: 基础目录路径 |
| |
| Returns: |
| (actual_sample_rate, needs_resample): 实际采样率和是否需要重采样 |
| """ |
| if pd.isna(audio_path) or not audio_path or not audio_path.strip(): |
| return None, False |
| |
| |
| full_path = os.path.join(base_dir, audio_path) |
| |
| if not os.path.exists(full_path): |
| print(f"警告: 音频文件不存在: {full_path}") |
| return None, False |
| |
| try: |
| |
| info = sf.info(full_path) |
| actual_sample_rate = info.samplerate |
| |
| |
| if actual_sample_rate != target_sample_rate: |
| |
| print(f"重采样: {audio_path} ({actual_sample_rate}Hz -> {target_sample_rate}Hz)") |
| |
| |
| waveform, sample_rate = torchaudio.load(full_path) |
| |
| |
| if waveform.shape[0] > 1: |
| waveform = torch.mean(waveform, dim=0, keepdim=True) |
| |
| |
| resampler = torchaudio.transforms.Resample(sample_rate, target_sample_rate) |
| waveform_resampled = resampler(waveform) |
| |
| |
| waveform_np = waveform_resampled.squeeze().numpy() |
| waveform_np = waveform_np.clip(-1.0, 1.0) |
| |
| |
| sf.write(full_path, waveform_np, target_sample_rate, subtype='PCM_16') |
| |
| return target_sample_rate, True |
| else: |
| |
| return actual_sample_rate, False |
| |
| except Exception as e: |
| print(f"错误: 处理音频文件 {full_path} 时出错: {e}") |
| return None, False |
|
|
| |
| print("检查并重采样query音频文件...") |
| query_sample_rates = [] |
| query_resampled_count = 0 |
|
|
| for idx, audio_path in enumerate(tqdm(df['query_audio_path'], desc="处理query音频")): |
| sample_rate, was_resampled = check_and_resample_audio(audio_path, base_dir) |
| query_sample_rates.append(sample_rate if sample_rate else target_sample_rate) |
| if was_resampled: |
| query_resampled_count += 1 |
|
|
| df['query_audio_sample_rate'] = query_sample_rates |
|
|
| |
| print("检查并重采样document音频文件...") |
| doc_sample_rates = [] |
| doc_resampled_count = 0 |
|
|
| for idx, audio_path in enumerate(tqdm(df['document_audio_path'], desc="处理document音频")): |
| sample_rate, was_resampled = check_and_resample_audio(audio_path, base_dir) |
| doc_sample_rates.append(sample_rate if sample_rate else target_sample_rate) |
| if was_resampled: |
| doc_resampled_count += 1 |
|
|
| df['document_audio_sample_rate'] = doc_sample_rates |
|
|
| |
| df.to_parquet('data.parquet', index=False) |
|
|
| print(f'\n完成!') |
| print(f'Query音频: 重采样了 {query_resampled_count}/{len(df)} 个文件') |
| print(f'Document音频: 重采样了 {doc_resampled_count}/{len(df)} 个文件') |
| print(f'新列: {df.columns.tolist()}') |
| print(f'形状: {df.shape}') |
| print(df.head(2)) |
|
|