Update utils/mir_eval_modules.py

utils/mir_eval_modules.py

@@ -28,23 +28,62 @@ def idx2voca_chord():
 
 def audio_file_to_features(audio_file, config):
     original_wav, sr = librosa.load(audio_file, sr=config.mp3['song_hz'], mono=True)
+
+    # Total duration in seconds
+    total_duration = librosa.get_duration(y=original_wav, sr=sr)
+
+    # Define segment length in seconds (you can still use config.mp3['inst_len'] for the segment size)
+    segment_len_sec = config.mp3.get('inst_len', 10)  # default to 10s if not defined
+
     currunt_sec_hz = 0
-    while len(original_wav) > currunt_sec_hz + config.mp3['song_hz'] * config.mp3['inst_len']:
+    feature = None
+
+    while currunt_sec_hz + sr * segment_len_sec < len(original_wav):
         start_idx = int(currunt_sec_hz)
-        end_idx = int(currunt_sec_hz + config.mp3['song_hz'] * config.mp3['inst_len'])
-        tmp = librosa.cqt(original_wav[start_idx:end_idx], sr=sr, n_bins=config.feature['n_bins'], bins_per_octave=config.feature['bins_per_octave'], hop_length=config.feature['hop_length'])
-        if start_idx == 0:
-            feature = tmp
-        else:
-            feature = np.concatenate((feature, tmp), axis=1)
+        end_idx = int(currunt_sec_hz + sr * segment_len_sec)
+        tmp = librosa.cqt(
+            original_wav[start_idx:end_idx], sr=sr,
+            n_bins=config.feature['n_bins'],
+            bins_per_octave=config.feature['bins_per_octave'],
+            hop_length=config.feature['hop_length']
+        )
+        feature = tmp if feature is None else np.concatenate((feature, tmp), axis=1)
         currunt_sec_hz = end_idx
-    tmp = librosa.cqt(original_wav[currunt_sec_hz:], sr=sr, n_bins=config.feature['n_bins'], bins_per_octave=config.feature['bins_per_octave'], hop_length=config.feature['hop_length'])
+
+    # Handle the final chunk (remaining audio)
+    tmp = librosa.cqt(
+        original_wav[currunt_sec_hz:], sr=sr,
+        n_bins=config.feature['n_bins'],
+        bins_per_octave=config.feature['bins_per_octave'],
+        hop_length=config.feature['hop_length']
+    )
     feature = np.concatenate((feature, tmp), axis=1)
     feature = np.log(np.abs(feature) + 1e-6)
-    feature_per_second = config.mp3['inst_len'] / config.model['timestep']
-    song_length_second = len(original_wav)/config.mp3['song_hz']
+
+    feature_per_second = segment_len_sec / config.model['timestep']
+    song_length_second = len(original_wav) / sr
+
     return feature, feature_per_second, song_length_second
 
+# def audio_file_to_features(audio_file, config):
+#     original_wav, sr = librosa.load(audio_file, sr=config.mp3['song_hz'], mono=True)
+#     currunt_sec_hz = 0
+#     while len(original_wav) > currunt_sec_hz + config.mp3['song_hz'] * config.mp3['inst_len']:
+#         start_idx = int(currunt_sec_hz)
+#         end_idx = int(currunt_sec_hz + config.mp3['song_hz'] * config.mp3['inst_len'])
+#         tmp = librosa.cqt(original_wav[start_idx:end_idx], sr=sr, n_bins=config.feature['n_bins'], bins_per_octave=config.feature['bins_per_octave'], hop_length=config.feature['hop_length'])
+#         if start_idx == 0:
+#             feature = tmp
+#         else:
+#             feature = np.concatenate((feature, tmp), axis=1)
+#         currunt_sec_hz = end_idx
+#     tmp = librosa.cqt(original_wav[currunt_sec_hz:], sr=sr, n_bins=config.feature['n_bins'], bins_per_octave=config.feature['bins_per_octave'], hop_length=config.feature['hop_length'])
+#     feature = np.concatenate((feature, tmp), axis=1)
+#     feature = np.log(np.abs(feature) + 1e-6)
+#     feature_per_second = config.mp3['inst_len'] / config.model['timestep']
+#     song_length_second = len(original_wav)/config.mp3['song_hz']
+#     return feature, feature_per_second, song_length_second
+
 # Audio files with format of wav and mp3
 def get_audio_paths(audio_dir):
     return [os.path.join(root, fname) for (root, dir_names, file_names) in os.walk(audio_dir, followlinks=True)