From 658f85f18fd676830384f491e5206a7632abca99 Mon Sep 17 00:00:00 2001
From: Lucasmotabr <lucasmotabr@users.noreply.github.com>
Date: Mon, 13 Apr 2026 16:17:38 +0900
Subject: [PATCH] feat: add SoundDetection msg and ListenForSound action for
 yamnet_ros

Adds the two interfaces required by the yamnet_ros bell/sound detection
package: SoundDetection.msg (detection event topic) and
ListenForSound.action (blocking action server for SMACH integration).
Removes ModeCtrl.srv which is no longer used by the team.
---
 CMakeLists.txt               |  2 ++
 action/ListenForSound.action | 14 ++++++++++++++
 msg/SoundDetection.msg       |  5 +++++
 3 files changed, 21 insertions(+)
 create mode 100644 action/ListenForSound.action
 create mode 100644 msg/SoundDetection.msg

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 478a607..139af40 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -19,6 +19,7 @@ set(msg_files
   msg/Feature.msg
   msg/DetectMask.msg
   msg/DetectMaskArray.msg
+  msg/SoundDetection.msg
 )
 set(srv_files
   srv/GetHandToTargetCoord.srv
@@ -39,6 +40,7 @@ set(action_files
   action/MoveWheelRotate.action
   action/VlaRecordState.action
   action/DisplayControl.action
+  action/ListenForSound.action
 )
 
 rosidl_generate_interfaces(${PROJECT_NAME}
diff --git a/action/ListenForSound.action b/action/ListenForSound.action
new file mode 100644
index 0000000..c06fd2b
--- /dev/null
+++ b/action/ListenForSound.action
@@ -0,0 +1,14 @@
+# Goal — start listening for a bell/doorbell sound
+float64 timeout_sec     # max seconds to listen; 0 = no timeout
+float32 threshold 0.15  # YAMNet score threshold to trigger; 0 = use node default
+---
+# Result
+bool    detected        # true if sound was detected before timeout
+string  label           # class name that triggered (e.g. "Doorbell")
+float32 score           # score of the detected class
+float32 elapsed_time    # seconds from goal start to result
+---
+# Feedback — sent every hop (~0.5 s) while listening
+string  current_top_label    # overall top class right now
+float32 current_top_score    # its score
+bool    candidate_detected   # any doorbell-like class above threshold this frame
diff --git a/msg/SoundDetection.msg b/msg/SoundDetection.msg
new file mode 100644
index 0000000..5c1ce95
--- /dev/null
+++ b/msg/SoundDetection.msg
@@ -0,0 +1,5 @@
+std_msgs/Header header
+string label        # best doorbell-like class detected (e.g. "Doorbell", "Bell")
+float32 score       # YAMNet score for that class [0-1]
+string top_label    # overall top class regardless of category (for diagnostics)
+float32 top_score   # overall top score (for diagnostics)