'update'

k2-fsa · Nov 27, 2024 · 51efa14 · 51efa14
1 parent cbe012d
commit 51efa14
Show file tree

Hide file tree

Showing 8 changed files with 74 additions and 2 deletions.
diff --git a/.idea/.gitignore b/.idea/.gitignore
diff --git a/.idea/icefall.iml b/.idea/icefall.iml
diff --git a/.idea/inspectionProfiles/Project_Default.xml b/.idea/inspectionProfiles/Project_Default.xml
diff --git a/.idea/inspectionProfiles/profiles_settings.xml b/.idea/inspectionProfiles/profiles_settings.xml
diff --git a/.idea/modules.xml b/.idea/modules.xml
diff --git a/.idea/vcs.xml b/.idea/vcs.xml
diff --git a/egs/multi_zh-hans/ASR/local/compute_fbank_kespeech_dev_test.py b/egs/multi_zh-hans/ASR/local/compute_fbank_kespeech_dev_test.py
@@ -52,7 +52,13 @@ def get_parser():
         default=80,
         help="""The number of mel bins for Fbank""",
     )
-
+    # 修改： 添加 指定参数 speed-perturb
+    parser.add_argument(
+        "--speed-perturb",
+        type=bool,
+        default=False,
+        help="Enable 0.9 and 1.1 speed perturbation for data augmentation. Default: False.",
+    )
     parser.add_argument(
         "--whisper-fbank",
         type=str2bool,
@@ -103,7 +109,11 @@ def compute_fbank_kespeech_dev_test(args):
         cut_set = cut_set.trim_to_supervisions(
             keep_overlapping=False, min_duration=None
         )
-
+        # 修改 执行 perturb操作
+        if speed_perturb:
+            cut_set = (
+                    cut_set + cut_set.perturb_speed(0.9) + cut_set.perturb_speed(1.1)
+            )
         logging.info("Computing features")
         cut_set = cut_set.compute_and_store_features_batch(
             extractor=extractor,

diff --git a/egs/multi_zh-hans/ASR/local/compute_fbank_kespeech_splits.py b/egs/multi_zh-hans/ASR/local/compute_fbank_kespeech_splits.py
@@ -100,6 +100,19 @@ def get_parser():
         help="""The number of mel bins for Fbank""",
     )
 
+    parser.add_argument(
+        "--whisper-fbank",
+        type=str2bool,
+        default=False,
+        help="Use WhisperFbank instead of Fbank. Default: False.",
+    )
+    #修改： 添加 指定参数 speed-perturb
+    parser.add_argument(
+        "--speed-perturb",
+        type=bool,
+        default=False,
+        help="Enable 0.9 and 1.1 speed perturbation for data augmentation. Default: False.",
+    )
     parser.add_argument(
         "--whisper-fbank",
         type=str2bool,
@@ -157,6 +170,11 @@ def compute_fbank_kespeech_splits(args):
         cut_set = cut_set.trim_to_supervisions(
             keep_overlapping=False, min_duration=None
         )
+        # 修改 执行 perturb操作
+        if speed_perturb:
+            cut_set = (
+                    cut_set + cut_set.perturb_speed(0.9) + cut_set.perturb_speed(1.1)
+            )
 
         logging.info("Computing features")
         cut_set = cut_set.compute_and_store_features_batch(