github actions

.github/workflows/black.yml

@@ -13,5 +13,5 @@ jobs:
       - uses: actions/checkout@v2
       - uses: psf/black@stable
         with:
-          options: "--check --diff --color"
-          version: "21.10b0"
+          options: "--check --diff --color --extend-exclude '/hyp_utils\\/kaldi/'"
+          version: "21.10b0"

egs/chime5_spkdet/v1/local/make_musan.py

@@ -7,125 +7,183 @@
 
 import os, sys
 
+
 def process_music_annotations(path):
-  utt2spk = {}
-  utt2vocals = {}
-  lines = open(path, 'r').readlines()
-  for line in lines:
-    utt, genres, vocals, musician = line.rstrip().split()[:4]
-    # For this application, the musican ID isn't important
-    utt2spk[utt] = utt
-    utt2vocals[utt] = vocals == "Y"
-  return utt2spk, utt2vocals
+    utt2spk = {}
+    utt2vocals = {}
+    lines = open(path, "r").readlines()
+    for line in lines:
+        utt, genres, vocals, musician = line.rstrip().split()[:4]
+        # For this application, the musican ID isn't important
+        utt2spk[utt] = utt
+        utt2vocals[utt] = vocals == "Y"
+    return utt2spk, utt2vocals
 
 
 def prepare_music(root_dir, fs, use_vocals):
-  utt2vocals = {}
-  utt2spk = {}
-  utt2wav = {}
-  num_good_files = 0
-  num_bad_files = 0
-  music_dir = os.path.join(root_dir, "music")
-  for root, dirs, files in os.walk(music_dir):
-    for file in files:
-      file_path = os.path.join(root, file)
-      if file.endswith(".wav"):
-        utt = str(file).replace(".wav", "")
-        utt2wav[utt] = file_path
-      elif str(file) == "ANNOTATIONS":
-        utt2spk_part, utt2vocals_part = process_music_annotations(file_path)
-        utt2spk.update(utt2spk_part)
-        utt2vocals.update(utt2vocals_part)
-  utt2spk_str = ""
-  utt2wav_str = ""
-  for utt in utt2vocals:
-    if utt in utt2wav:
-      if use_vocals or not utt2vocals[utt]:
-        utt2spk_str = utt2spk_str + utt + " " + utt2spk[utt] + "\n"
-        if fs == 8:
-          utt2wav_str = utt2wav_str + utt + " sox -t wav " + utt2wav[utt] + " -r 8k -t wav - |\n"
+    utt2vocals = {}
+    utt2spk = {}
+    utt2wav = {}
+    num_good_files = 0
+    num_bad_files = 0
+    music_dir = os.path.join(root_dir, "music")
+    for root, dirs, files in os.walk(music_dir):
+        for file in files:
+            file_path = os.path.join(root, file)
+            if file.endswith(".wav"):
+                utt = str(file).replace(".wav", "")
+                utt2wav[utt] = file_path
+            elif str(file) == "ANNOTATIONS":
+                utt2spk_part, utt2vocals_part = process_music_annotations(file_path)
+                utt2spk.update(utt2spk_part)
+                utt2vocals.update(utt2vocals_part)
+    utt2spk_str = ""
+    utt2wav_str = ""
+    for utt in utt2vocals:
+        if utt in utt2wav:
+            if use_vocals or not utt2vocals[utt]:
+                utt2spk_str = utt2spk_str + utt + " " + utt2spk[utt] + "\n"
+                if fs == 8:
+                    utt2wav_str = (
+                        utt2wav_str
+                        + utt
+                        + " sox -t wav "
+                        + utt2wav[utt]
+                        + " -r 8k -t wav - |\n"
+                    )
+                else:
+                    utt2wav_str = (
+                        utt2wav_str
+                        + utt
+                        + " sox -t wav "
+                        + utt2wav[utt]
+                        + " -r 16k -t wav - |\n"
+                    )
+            num_good_files += 1
         else:
-          utt2wav_str = utt2wav_str + utt + " sox -t wav " + utt2wav[utt] + " -r 16k -t wav - |\n"
-      num_good_files += 1
-    else:
-      print("Missing file", utt)
-      num_bad_files += 1
-  print("In music directory, processed", num_good_files, "files;", num_bad_files, "had missing wav data")
-  return utt2spk_str, utt2wav_str
+            print("Missing file", utt)
+            num_bad_files += 1
+    print(
+        "In music directory, processed",
+        num_good_files,
+        "files;",
+        num_bad_files,
+        "had missing wav data",
+    )
+    return utt2spk_str, utt2wav_str
+
 
 def prepare_speech(root_dir, fs):
-  utt2spk = {}
-  utt2wav = {}
-  num_good_files = 0
-  num_bad_files = 0
-  speech_dir = os.path.join(root_dir, "speech")
-  for root, dirs, files in os.walk(speech_dir):
-    for file in files:
-      file_path = os.path.join(root, file)
-      if file.endswith(".wav"):
-        utt = str(file).replace(".wav", "")
-        utt2wav[utt] = file_path
-        utt2spk[utt] = utt
-  utt2spk_str = ""
-  utt2wav_str = ""
-  for utt in utt2spk:
-    if utt in utt2wav:
-      utt2spk_str = utt2spk_str + utt + " " + utt2spk[utt] + "\n"
-      if fs == 8:
-        utt2wav_str = utt2wav_str + utt + " sox -t wav " + utt2wav[utt] + " -r 8k -t wav - |\n"
-      else:
-        utt2wav_str = utt2wav_str + utt + " sox -t wav " + utt2wav[utt] + " -r 16k -t wav - |\n"
-      num_good_files += 1
-    else:
-      print("Missing file", utt)
-      num_bad_files += 1
-  print("In speech directory, processed", num_good_files, "files;", num_bad_files, "had missing wav data")
-  return utt2spk_str, utt2wav_str
+    utt2spk = {}
+    utt2wav = {}
+    num_good_files = 0
+    num_bad_files = 0
+    speech_dir = os.path.join(root_dir, "speech")
+    for root, dirs, files in os.walk(speech_dir):
+        for file in files:
+            file_path = os.path.join(root, file)
+            if file.endswith(".wav"):
+                utt = str(file).replace(".wav", "")
+                utt2wav[utt] = file_path
+                utt2spk[utt] = utt
+    utt2spk_str = ""
+    utt2wav_str = ""
+    for utt in utt2spk:
+        if utt in utt2wav:
+            utt2spk_str = utt2spk_str + utt + " " + utt2spk[utt] + "\n"
+            if fs == 8:
+                utt2wav_str = (
+                    utt2wav_str
+                    + utt
+                    + " sox -t wav "
+                    + utt2wav[utt]
+                    + " -r 8k -t wav - |\n"
+                )
+            else:
+                utt2wav_str = (
+                    utt2wav_str
+                    + utt
+                    + " sox -t wav "
+                    + utt2wav[utt]
+                    + " -r 16k -t wav - |\n"
+                )
+            num_good_files += 1
+        else:
+            print("Missing file", utt)
+            num_bad_files += 1
+    print(
+        "In speech directory, processed",
+        num_good_files,
+        "files;",
+        num_bad_files,
+        "had missing wav data",
+    )
+    return utt2spk_str, utt2wav_str
+
 
 def prepare_noise(root_dir, fs):
-  utt2spk = {}
-  utt2wav = {}
-  num_good_files = 0
-  num_bad_files = 0
-  noise_dir = os.path.join(root_dir, "noise")
-  for root, dirs, files in os.walk(noise_dir):
-    for file in files:
-      file_path = os.path.join(root, file)
-      if file.endswith(".wav"):
-        utt = str(file).replace(".wav", "")
-        utt2wav[utt] = file_path
-        utt2spk[utt] = utt
-  utt2spk_str = ""
-  utt2wav_str = ""
-  for utt in utt2spk:
-    if utt in utt2wav:
-      utt2spk_str = utt2spk_str + utt + " " + utt2spk[utt] + "\n"
-      if fs == 8:
-        utt2wav_str = utt2wav_str + utt + " sox -t wav " + utt2wav[utt] + " -r 8k -t wav - |\n"
-      else:
-        utt2wav_str = utt2wav_str + utt + " sox -t wav " + utt2wav[utt] + " -r 16k -t wav - |\n"
-      num_good_files += 1
-    else:
-      print("Missing file", utt)
-      num_bad_files += 1
-  print("In noise directory, processed", num_good_files, "files;", num_bad_files, "had missing wav data")
-  return utt2spk_str, utt2wav_str
+    utt2spk = {}
+    utt2wav = {}
+    num_good_files = 0
+    num_bad_files = 0
+    noise_dir = os.path.join(root_dir, "noise")
+    for root, dirs, files in os.walk(noise_dir):
+        for file in files:
+            file_path = os.path.join(root, file)
+            if file.endswith(".wav"):
+                utt = str(file).replace(".wav", "")
+                utt2wav[utt] = file_path
+                utt2spk[utt] = utt
+    utt2spk_str = ""
+    utt2wav_str = ""
+    for utt in utt2spk:
+        if utt in utt2wav:
+            utt2spk_str = utt2spk_str + utt + " " + utt2spk[utt] + "\n"
+            if fs == 8:
+                utt2wav_str = (
+                    utt2wav_str
+                    + utt
+                    + " sox -t wav "
+                    + utt2wav[utt]
+                    + " -r 8k -t wav - |\n"
+                )
+            else:
+                utt2wav_str = (
+                    utt2wav_str
+                    + utt
+                    + " sox -t wav "
+                    + utt2wav[utt]
+                    + " -r 16k -t wav - |\n"
+                )
+            num_good_files += 1
+        else:
+            print("Missing file", utt)
+            num_bad_files += 1
+    print(
+        "In noise directory, processed",
+        num_good_files,
+        "files;",
+        num_bad_files,
+        "had missing wav data",
+    )
+    return utt2spk_str, utt2wav_str
+
 
 def main():
-  in_dir = sys.argv[1]
-  fs = int(sys.argv[2])
-  out_dir = sys.argv[3]
-  use_vocals = sys.argv[4] == "Y"
-  utt2spk_music, utt2wav_music = prepare_music(in_dir, fs, use_vocals)
-  utt2spk_speech, utt2wav_speech = prepare_speech(in_dir, fs)
-  utt2spk_noise, utt2wav_noise = prepare_noise(in_dir, fs)
-  utt2spk = utt2spk_speech + utt2spk_music + utt2spk_noise
-  utt2wav = utt2wav_speech + utt2wav_music + utt2wav_noise
-  wav_fi = open(os.path.join(out_dir, "wav.scp"), 'w')
-  wav_fi.write(utt2wav)
-  utt2spk_fi = open(os.path.join(out_dir, "utt2spk"), 'w')
-  utt2spk_fi.write(utt2spk)
+    in_dir = sys.argv[1]
+    fs = int(sys.argv[2])
+    out_dir = sys.argv[3]
+    use_vocals = sys.argv[4] == "Y"
+    utt2spk_music, utt2wav_music = prepare_music(in_dir, fs, use_vocals)
+    utt2spk_speech, utt2wav_speech = prepare_speech(in_dir, fs)
+    utt2spk_noise, utt2wav_noise = prepare_noise(in_dir, fs)
+    utt2spk = utt2spk_speech + utt2spk_music + utt2spk_noise
+    utt2wav = utt2wav_speech + utt2wav_music + utt2wav_noise
+    wav_fi = open(os.path.join(out_dir, "wav.scp"), "w")
+    wav_fi.write(utt2wav)
+    utt2spk_fi = open(os.path.join(out_dir, "utt2spk"), "w")
+    utt2spk_fi.write(utt2spk)
 
 
-if __name__=="__main__":
-  main()
+if __name__ == "__main__":
+    main()

egs/chime5_spkdet/v1/local/score_dcf.py

@@ -20,48 +20,54 @@
 
 def score_dcf(key_file, score_file, output_path):
 
-    logging.info('Load key: %s' % key_file)
+    logging.info("Load key: %s" % key_file)
     key = TrialKey.load_txt(key_file)
-    logging.info('Load scores: %s' % score_file)
+    logging.info("Load scores: %s" % score_file)
     scr = TrialScores.load_txt(score_file)
     tar, non = scr.get_tar_non(key)
 
-    priors = np.array([0.001, 0.005, 0.01, 0.05 ])
+    priors = np.array([0.001, 0.005, 0.01, 0.05])
     min_dcf, act_dcf, eer, _ = fast_eval(tar, non, priors)
-    
+
     output_dir = os.path.dirname(output_path)
     if not os.path.isdir(output_dir):
         os.makedirs(output_dir)
 
-    output_file = output_path + '_results'
-    with open(output_file, 'w') as f:
-        s = 'EER: {0:.2f} DCF5e-2: {1:.3f} / {2:.3f} DCF1e-2: {3:.3f} / {4:.3f} DCF5e-3: {5:.3f} / {6:.3f} DCF1e-3: {7:.3f} / {8:.3f}'.format(
-            eer * 100, min_dcf[3], act_dcf[3],
-            min_dcf[2], act_dcf[2],
-            min_dcf[1], act_dcf[1],
-            min_dcf[0], act_dcf[0])
+    output_file = output_path + "_results"
+    with open(output_file, "w") as f:
+        s = "EER: {0:.2f} DCF5e-2: {1:.3f} / {2:.3f} DCF1e-2: {3:.3f} / {4:.3f} DCF5e-3: {5:.3f} / {6:.3f} DCF1e-3: {7:.3f} / {8:.3f}".format(
+            eer * 100,
+            min_dcf[3],
+            act_dcf[3],
+            min_dcf[2],
+            act_dcf[2],
+            min_dcf[1],
+            act_dcf[1],
+            min_dcf[0],
+            act_dcf[0],
+        )
         f.write(s)
         logging.info(s)
-        
+
 
 if __name__ == "__main__":
 
-    parser=argparse.ArgumentParser(
-        formatter_class=argparse.ArgumentDefaultsHelpFormatter,                
-        fromfile_prefix_chars='@',
-        description='Computes EER and DCF')
+    parser = argparse.ArgumentParser(
+        formatter_class=argparse.ArgumentDefaultsHelpFormatter,
+        fromfile_prefix_chars="@",
+        description="Computes EER and DCF",
+    )
+
+    parser.add_argument("--key-file", dest="key_file", required=True)
+    parser.add_argument("--score-file", dest="score_file", required=True)
+    parser.add_argument("--output-path", dest="output_path", required=True)
+    parser.add_argument(
+        "-v", "--verbose", dest="verbose", default=1, choices=[0, 1, 2, 3], type=int
+    )
 
-    parser.add_argument('--key-file', dest='key_file', required=True)
-    parser.add_argument('--score-file', dest='score_file', required=True)
-    parser.add_argument('--output-path', dest='output_path', required=True)
-    parser.add_argument('-v', '--verbose', dest='verbose', default=1,
-                        choices=[0, 1, 2, 3], type=int)
-
-    args=parser.parse_args()
+    args = parser.parse_args()
     config_logger(args.verbose)
     del args.verbose
     logging.debug(args)
-
-    score_dcf(**vars(args))
-
 
+    score_dcf(**vars(args))