Summary

oceanai/__init__.py

@@ -18,7 +18,7 @@
 __uri__ = "https://github.com/DmitryRyumin/oceanai"
 
 __version__ = "1.0"
-__release__ = __version__ + ".0a20"
+__release__ = __version__ + ".0a21"
 
 __author__ru__ = "Рюмина Елена, Рюмин Дмитрий, Карпов Алексей"
 __author__en__ = "Elena Ryumina, Dmitry Ryumin, Alexey Karpov"
@@ -29,4 +29,4 @@
 __maintainer_email__ = "ryumina_ev@mail.ru, dl_03.03.1991@mail.ru"
 
 __license__ = "BSD License"
-__copyright__ = "2022-23. Speech and Multimodal Interfaces Laboratory Team"
+__copyright__ = "2022-24. Speech and Multimodal Interfaces Laboratory Team"

oceanai/modules/lab/audio.py

@@ -26,6 +26,7 @@
 import librosa  # Обработка аудио
 import audioread  # Декодирование звука
 import math
+import gradio
 
 from urllib.parse import urlparse
 from urllib.error import URLError
@@ -1181,8 +1182,7 @@ def _get_acoustic_features(
         try:
             # Проверка аргументов
             if (
-                type(path) is not str
-                or not path
+                (type(path) is not str or not path) and (type(path) is not gradio.utils.NamedString)
                 or type(sr) is not int
                 or sr < 1
                 or ((type(window) is not int or window < 1) and (type(window) is not float or window <= 0))

oceanai/modules/lab/keras_vggface/models.py

@@ -38,15 +38,15 @@ def resnet_identity_block(input_tensor, kernel_size, filters, stage, block, bias
     conv3_name = "conv" + str(stage) + "_" + str(block) + "_3x3"
 
     x = Conv2D(filters1, (1, 1), use_bias=bias, name=conv1_reduce_name)(input_tensor)
-    x = BatchNormalization(axis=bn_axis, name=conv1_reduce_name + "/bn")(x)
+    x = BatchNormalization(axis=bn_axis, name=conv1_reduce_name + "_bn")(x)
     x = Activation("relu")(x)
 
     x = Conv2D(filters2, kernel_size, use_bias=bias, padding="same", name=conv3_name)(x)
-    x = BatchNormalization(axis=bn_axis, name=conv3_name + "/bn")(x)
+    x = BatchNormalization(axis=bn_axis, name=conv3_name + "_bn")(x)
     x = Activation("relu")(x)
 
     x = Conv2D(filters3, (1, 1), use_bias=bias, name=conv1_increase_name)(x)
-    x = BatchNormalization(axis=bn_axis, name=conv1_increase_name + "/bn")(x)
+    x = BatchNormalization(axis=bn_axis, name=conv1_increase_name + "_bn")(x)
 
     x = layers.add([x, input_tensor])
     x = Activation("relu")(x)
@@ -65,18 +65,18 @@ def resnet_conv_block(input_tensor, kernel_size, filters, stage, block, strides=
     conv3_name = "conv" + str(stage) + "_" + str(block) + "_3x3"
 
     x = Conv2D(filters1, (1, 1), strides=strides, use_bias=bias, name=conv1_reduce_name)(input_tensor)
-    x = BatchNormalization(axis=bn_axis, name=conv1_reduce_name + "/bn")(x)
+    x = BatchNormalization(axis=bn_axis, name=conv1_reduce_name + "_bn")(x)
     x = Activation("relu")(x)
 
     x = Conv2D(filters2, kernel_size, padding="same", use_bias=bias, name=conv3_name)(x)
-    x = BatchNormalization(axis=bn_axis, name=conv3_name + "/bn")(x)
+    x = BatchNormalization(axis=bn_axis, name=conv3_name + "_bn")(x)
     x = Activation("relu")(x)
 
     x = Conv2D(filters3, (1, 1), name=conv1_increase_name, use_bias=bias)(x)
-    x = BatchNormalization(axis=bn_axis, name=conv1_increase_name + "/bn")(x)
+    x = BatchNormalization(axis=bn_axis, name=conv1_increase_name + "_bn")(x)
 
     shortcut = Conv2D(filters3, (1, 1), strides=strides, use_bias=bias, name=conv1_proj_name)(input_tensor)
-    shortcut = BatchNormalization(axis=bn_axis, name=conv1_proj_name + "/bn")(shortcut)
+    shortcut = BatchNormalization(axis=bn_axis, name=conv1_proj_name + "_bn")(shortcut)
 
     x = layers.add([x, shortcut])
     x = Activation("relu")(x)
@@ -100,8 +100,8 @@ def RESNET50(include_top=True, weights="vggface", input_shape=None, pooling=None
     else:
         bn_axis = 1
 
-    x = Conv2D(64, (7, 7), use_bias=False, strides=(2, 2), padding="same", name="conv1/7x7_s2")(img_input)
-    x = BatchNormalization(axis=bn_axis, name="conv1/7x7_s2/bn")(x)
+    x = Conv2D(64, (7, 7), use_bias=False, strides=(2, 2), padding="same", name="conv1_7x7_s2")(img_input)
+    x = BatchNormalization(axis=bn_axis, name="conv1_7x7_s2_bn")(x)
     x = Activation("relu")(x)
     x = MaxPooling2D((3, 3), strides=(2, 2))(x)