Add paramter for detection threshold

launch/singleview_loop.launch

 <launch>
   <arg name="dataset" default="ycbv"/>
+  <arg name="detection_threshold" default="0.7"/>
   <arg name="debug" default="false"/>
   <arg name="debug_models" default="bop_datasets/$(arg dataset)/models/"/> <!-- Relative path from cosypose local_data directory -->
 
@@ -12,6 +13,7 @@
   <!-- Start service and camera -->
   <include file="$(find ros_cosypose)/launch/singleview_service.launch">
     <arg name="dataset" value="$(arg dataset)"/>
+    <arg name="detection_threshold" value="$(arg detection_threshold)"/>
     <arg name="debug" value="$(arg debug)"/>
     <arg name="debug_models" value="$(arg debug_models)"/>
 
@@ -21,7 +23,7 @@
   </include>
 
   <!-- call ros_cosypose in loop and publish on topics -->
-  <node name="$(anon pose_estimation_loop.py)" pkg="ros_cosypose" type="pose_estimation_loop.py">
+  <node name="pose_estimation_loop" pkg="ros_cosypose" type="pose_estimation_loop.py">
     <param name="camera_name" value="$(arg camera_name)"/>
     <param name="allow_tracking" value="$(arg allow_tracking)"/>
   </node>

launch/singleview_service.launch

 <launch>
     <arg name="dataset" default="ycbv"/>
+    <arg name="detection_threshold" default="0.7"/>
     <arg name="debug" default="false"/>
     <arg name="debug_models" default="bop_datasets/$(arg dataset)/models/"/> <!-- Relative path from cosypose local_data directory -->
 
@@ -18,8 +19,9 @@
     </group>
 
     <!-- ros_cosypose service -->
-    <node name="$(anon pose_estimation.py)" pkg="ros_cosypose" type="pose_estimation.py">
+    <node name="pose_estimation" pkg="ros_cosypose" type="pose_estimation.py">
       <param name="dataset" value="$(arg dataset)"/>
+      <param name="detection_threshold" value="$(arg detection_threshold)"/>
       <param name="debug" value="$(arg debug)"/>
       <param name="debug_models" value="$(arg debug_models)"/>
     </node>

scripts/pose_estimation.py

@@ -39,11 +39,6 @@ class PredictionService:
             object_detector_run_id=object_detector_run_id,
         )
 
-        self.detector_kwargs = dict(
-            one_instance_per_class=False,
-            detection_th=0.9
-        )
-
         self.debug_converter = None
         if(debug):
             self.debug_converter = RVizTranslator(debug_models)
@@ -79,8 +74,10 @@ class PredictionService:
             pose_estimation_prior = PandasTensorCollection(infos=pd.DataFrame(dict(label=labels,batch_im_id=batch_im_ids)),
                                                            poses=torch.tensor(poses).float().cuda())
 
+        detector_kwargs = dict(one_instance_per_class=False, detection_th=rospy.get_param('~detection_threshold'))
+
         # Predict poses using cosypose
-        self.predictor([image, ], cameras, pose_estimation_prior=pose_estimation_prior, detector_kwargs=self.detector_kwargs)
+        self.predictor([image, ], cameras, pose_estimation_prior=pose_estimation_prior, detector_kwargs=detector_kwargs)
         pose_predictions = self.predictor.pose_predictions
 
         # transform cosypose objects to ros msg
@@ -107,6 +104,9 @@ if __name__ == '__main__':
     debug = rospy.get_param('~debug', "false")
     debug_models = rospy.get_param('~debug_models', "")
 
+    if not rospy.has_param('~detection_threshold'):
+        rospy.set_param('~detection_threshold', 0.7)
+
     prediction_srv = PredictionService(dataset, debug, debug_models)
 
     rospy.Service('pose_estimation', GetPrediction, prediction_srv.pose_estimation)