diff --git a/projects/opendr_ws_2/src/opendr_interface/CMakeLists.txt b/projects/opendr_ws_2/src/opendr_interface/CMakeLists.txt
index 9c158812e5..21356f6664 100644
--- a/projects/opendr_ws_2/src/opendr_interface/CMakeLists.txt
+++ b/projects/opendr_ws_2/src/opendr_interface/CMakeLists.txt
@@ -31,6 +31,8 @@ rosidl_generate_interfaces(${PROJECT_NAME}
   "msg/OpenDRPose2DKeypoint.msg"
   "msg/OpenDRPose3D.msg"
   "msg/OpenDRPose3DKeypoint.msg"
+  "msg/AudioData.msg"
+  "msg/AudioInfo.msg"
   "srv/OpenDRSingleObjectTracking.srv"
   "srv/ImgToMesh.srv"
   DEPENDENCIES std_msgs shape_msgs sensor_msgs vision_msgs
diff --git a/projects/opendr_ws_2/src/opendr_interface/msg/AudioData.msg b/projects/opendr_ws_2/src/opendr_interface/msg/AudioData.msg
new file mode 100644
index 0000000000..dd2331bbe6
--- /dev/null
+++ b/projects/opendr_ws_2/src/opendr_interface/msg/AudioData.msg
@@ -0,0 +1 @@
+uint8[] data
diff --git a/projects/opendr_ws_2/src/opendr_interface/msg/AudioInfo.msg b/projects/opendr_ws_2/src/opendr_interface/msg/AudioInfo.msg
new file mode 100644
index 0000000000..ca4b55e34f
--- /dev/null
+++ b/projects/opendr_ws_2/src/opendr_interface/msg/AudioInfo.msg
@@ -0,0 +1,12 @@
+# This message contains the audio meta data
+
+# Number of channels
+uint8 channels
+# Sampling rate [Hz]
+uint32 sample_rate
+# Audio format (e.g. S16LE)
+string sample_format
+# Amount of audio data per second [bits/s]
+uint32 bitrate
+# Audio coding format (e.g. WAVE, MP3)
+string coding_format
diff --git a/projects/opendr_ws_2/src/opendr_perception/opendr_perception/audiovisual_emotion_recognition_node.py b/projects/opendr_ws_2/src/opendr_perception/opendr_perception/audiovisual_emotion_recognition_node.py
index 008b51d7b7..78c1b09629 100644
--- a/projects/opendr_ws_2/src/opendr_perception/opendr_perception/audiovisual_emotion_recognition_node.py
+++ b/projects/opendr_ws_2/src/opendr_perception/opendr_perception/audiovisual_emotion_recognition_node.py
@@ -25,7 +25,10 @@
 from rclpy.node import Node
 import message_filters
 from sensor_msgs.msg import Image as ROS_Image
-from audio_common_msgs.msg import AudioData
+if os.environ.get('ROS_DISTRO').lower() in ['foxy', 'humble']:
+  from opendr_interface.msg import AudioData
+else:
+  from audio_common_msgs.msg import AudioData
 from vision_msgs.msg import Classification2D
 
 from opendr_bridge import ROS2Bridge
diff --git a/projects/opendr_ws_2/src/opendr_perception/opendr_perception/speech_command_recognition_node.py b/projects/opendr_ws_2/src/opendr_perception/opendr_perception/speech_command_recognition_node.py
index 98344922bf..47a38b8d19 100755
--- a/projects/opendr_ws_2/src/opendr_perception/opendr_perception/speech_command_recognition_node.py
+++ b/projects/opendr_ws_2/src/opendr_perception/opendr_perception/speech_command_recognition_node.py
@@ -17,10 +17,14 @@
 import argparse
 import torch
 import numpy as np
+import os
 
 import rclpy
 from rclpy.node import Node
-from audio_common_msgs.msg import AudioData
+if os.environ.get('ROS_DISTRO').lower() in ['foxy', 'humble']:
+  from opendr_interface.msg import AudioData
+else:
+  from audio_common_msgs.msg import AudioData
 from vision_msgs.msg import Classification2D
 
 from opendr_bridge import ROS2Bridge
diff --git a/src/opendr/perception/facial_expression_recognition/landmark_based_facial_expression_recognition/progressive_spatio_temporal_bln_learner.py b/src/opendr/perception/facial_expression_recognition/landmark_based_facial_expression_recognition/progressive_spatio_temporal_bln_learner.py
index 0f591adc89..e75aeaa558 100644
--- a/src/opendr/perception/facial_expression_recognition/landmark_based_facial_expression_recognition/progressive_spatio_temporal_bln_learner.py
+++ b/src/opendr/perception/facial_expression_recognition/landmark_based_facial_expression_recognition/progressive_spatio_temporal_bln_learner.py
@@ -315,7 +315,7 @@ def eval(self, val_dataset, val_loader=None, epoch=0, monte_carlo_dropout=True,
         :type val_data_filename: str, optional
         :param val_labels_filename: the file name of val labels which is placed in the dataset path.
         :type val_labels_filename: str, optional
-        :param save_score: if set to True, it saves the classification score of all samples in differenc classes
+        :param save_score: if set to True, it saves the classification score of all samples in different classes
         in a log file. Default to False.
         :type save_score: bool, optional
         :param wrong_file: if set to True, it saves the results of wrongly classified samples. Default to False.