#!/usr/bin/env python3 """ 昔涟声纹搜索 — 多线程 + 全精度 pyin 用法: python search_cyrene.py 监控: tail -f cyrene_confirmed/search.log """ import os, sys, shutil, json, time, warnings, logging from multiprocessing import Pool, cpu_count import numpy as np import librosa warnings.filterwarnings('ignore') # ═══════════ 配置 ═══════════ SEARCH_DIR = r"D:\Project\Code\Uni\Cyrene-Voice-Model\data\cleaned" OUT_DIR = r"D:\Project\Code\Uni\Cyrene-Voice-Model\data\cyrene_confirmed" WORKERS = max(1, cpu_count() - 1) BATCH = 500 # 每批文件数 CONFIRMED = [ # ── tier1 确认 (22) ── "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0358_02ff22a9.wav", "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0369_0315d90a.wav", "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0370_0317b5bf.wav", "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0391_033cbeea.wav", "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0392_033d9e2e.wav", "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0398_0346fa30.wav", "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0417_0367cbc5.wav", "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0330_02cb864d.wav", "External_del_3.5_chapter_2/External_del_3.5_chapter_2_0357_02fd4ab1.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0106_009dfe25.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0107_009fba53.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0112_00a89eb2.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0124_00c08c68.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0128_00c6ba19.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0365_026cf0de.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0371_0278f79f.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0374_02803ab5.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0409_02bcd547.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0410_02bec9f4.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0437_02e57ff0.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0440_02e7d98f.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0467_031daa9e.wav", # ── tier2 新确认 (31) ── "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0102_0095ba46.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0109_00a458cc.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0126_00c44990.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0130_00c95298.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0359_02670f55.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0364_026b8023.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0366_026f3922.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0369_0274d62a.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0372_027b67ef.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0392_029e92a6.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0413_02c1a418.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0415_02c466f1.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0417_02c8ffc0.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0430_02dbad24.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0431_02dda165.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0434_02e216cf.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0443_02eb0bc4.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0455_03048430.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0462_0316bcd9.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0463_03181c56.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0464_03195142.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0468_031e8bd6.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0469_03208c8a.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0484_03339e50.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0488_033a7664.wav", "External_del_3.4_chapter_0/External_del_3.4_chapter_0_0489_033b5c44.wav", ] # ═══════════ 纯函数 (供 Pool 调用) ═══════════ def extract_features(wav_path): """全精度 pyin + MFCC20 + delta×2 + spectral""" try: y, sr = librosa.load(wav_path, sr=22050, mono=True) if len(y) < sr * 0.25: return None f0, _, _ = librosa.pyin(y, fmin=80, fmax=600, sr=sr) f0 = f0[~np.isnan(f0)] if len(f0) < 10: return None mfcc = librosa.feature.mfcc(y=y, sr=sr, n_mfcc=20) mfcc_d = librosa.feature.delta(mfcc) mfcc_d2 = librosa.feature.delta(mfcc, order=2) cent = librosa.feature.spectral_centroid(y=y, sr=sr) roll = librosa.feature.spectral_rolloff(y=y, sr=sr) return np.concatenate([ [np.mean(f0), np.std(f0), np.percentile(f0,10), np.percentile(f0,25), np.percentile(f0,50), np.percentile(f0,75), np.percentile(f0,90)], np.mean(mfcc,axis=1), np.std(mfcc,axis=1), np.mean(mfcc_d,axis=1), np.std(mfcc_d,axis=1), np.mean(mfcc_d2,axis=1), np.std(mfcc_d2,axis=1), [np.mean(cent), np.std(cent), np.mean(roll), np.std(roll)], ]).astype(np.float64) except: return None # ═══════════ 主流程 ═══════════ def main(): os.makedirs(OUT_DIR, exist_ok=True) LOG_FILE = os.path.join(OUT_DIR, "search.log") # 日志: 文件 + 终端 log = logging.getLogger("cyrene") log.setLevel(logging.INFO) for h in [logging.FileHandler(LOG_FILE, encoding='utf-8'), logging.StreamHandler(sys.stdout)]: h.setFormatter(logging.Formatter("%(asctime)s [%(levelname)s] %(message)s", "%H:%M:%S")) log.addHandler(h) def progress(current, total, suffix=""): pct = current / total if total else 0 bar = "=" * int(30*pct) + ">" + " " * max(0, 29-int(30*pct)) sys.stdout.write(f"\r [{bar}] {pct*100:5.1f}% {current}/{total} {suffix}") sys.stdout.flush() log.info("=" * 55) log.info("昔涟声纹搜索 — 多线程全精度模式") log.info(f" 线程: {WORKERS} | 批量: {BATCH}") log.info("=" * 55) t_start = time.time() # ── Step 1: 参考模板 ── log.info("\n[1/3] 提取参考模板...") ref_feats = [] for fname in CONFIRMED: path = os.path.join(SEARCH_DIR, fname) f = extract_features(path) if f is not None: ref_feats.append(f) log.info(f" OK {os.path.basename(fname)}") template = np.mean(ref_feats, axis=0) log.info(f" 模板: {len(ref_feats)} 文件 | pitch={template[0]:.0f}Hz | dim={len(template)}") # ── Step 2: 文件列表 ── log.info(f"\n[2/3] 收集文件...") all_wavs = [] for root, dirs, files in os.walk(SEARCH_DIR): for f in files: if f.endswith(".wav"): all_wavs.append(os.path.join(root, f)) log.info(f" {len(all_wavs):,} 个 WAV 文件") # ── Step 3: 多线程搜索 ── log.info(f"\n[3/3] 多线程搜索 ({len(all_wavs):,} 文件, {WORKERS} 线程)...") log.info("-" * 55) t2 = time.time() results = [] errors = 0 pool = Pool(WORKERS) done = 0 for feats in pool.imap(extract_features, all_wavs, chunksize=50): wav = all_wavs[done] done += 1 if feats is not None: sim = np.dot(feats, template) / (np.linalg.norm(feats) * np.linalg.norm(template) + 1e-8) p = feats[0] penalty = 1.0 / (1.0 + abs(p - template[0]) / 100) results.append((sim * 0.6 + penalty * 0.4, sim, p, wav)) else: errors += 1 # 进度更新 if done % 100 == 0 or done == len(all_wavs): elapsed = time.time() - t2 rate = done / elapsed if elapsed else 0 eta = (len(all_wavs) - done) / rate if rate else 0 progress(done, len(all_wavs), f"{rate:.0f}f/s ETA{eta:.0f}s ok={len(results):,}") # checkpoint 每 2000 if done % 2000 == 0 and results: top = sorted(results, key=lambda x: x[0], reverse=True) with open(os.path.join(OUT_DIR, f"ckpt_{done}.json"), 'w') as jf: json.dump([(float(x[0]), float(x[2]), x[3]) for x in top[:500]], jf) pool.close() pool.join() print() log.info(f" 完成: {len(results):,} ok | {errors} skip | {time.time()-t2:.0f}s ({len(results)/(time.time()-t2):.0f} f/s)") log.info(f" 总耗时: {time.time()-t_start:.0f}s") # ── 排序 ── results.sort(key=lambda x: x[0], reverse=True) log.info(f"\n{'='*55}") log.info(f"Top 50 候选 (★ = combo > 0.85)") log.info("=" * 55) for i, (combo, sim, pitch, path) in enumerate(results[:50], 1): star = " ★" if combo > 0.85 else "" d = os.path.basename(os.path.dirname(path)) f = os.path.basename(path) log.info(f" {i:2d}. [{combo:.4f}]{star} {d}/{f}") # ── 来源统计 ── log.info(f"\n来源分布 (combo > 0.85):") srcs = {} for combo, sim, pitch, path in results: if combo > 0.85: d = os.path.basename(os.path.dirname(path)) srcs[d] = srcs.get(d, 0) + 1 for d in sorted(srcs): log.info(f" {d}: {srcs[d]} 文件") # ── 导出 ── copied = 0 for combo, sim, pitch, path in results: if combo > 0.85: dst = os.path.join(OUT_DIR, os.path.basename(path)) if not os.path.exists(dst): shutil.copy2(path, dst) copied += 1 log.info(f"\n导出: {copied} 文件 → {OUT_DIR}") rp = os.path.join(OUT_DIR, "search_results.json") with open(rp, 'w') as f: json.dump([(float(s), float(p), w) for s, p, w in results], f, ensure_ascii=False) log.info(f"结果: {rp}") log.info(f"日志: {LOG_FILE}") log.info(f"\n{'='*55}") log.info("DONE") if __name__ == "__main__": main()