In the standalone version the tracker is currently not publishing any poses by default. This is mainly because in the standalone version I couldn't assume any tf to be present but at least the camera tf should be there. Therefore, I'll change the default behaviour to:
If there is no tf_target_frame specified, the pedestrian tracker will publish poses in the tf frame of the camera: *_depth_optical_frame
As mentioned in #89 this might be useful.
In the standalone version the tracker is currently not publishing any poses by default. This is mainly because in the standalone version I couldn't assume any tf to be present but at least the camera tf should be there. Therefore, I'll change the default behaviour to:
If there is no
tf_target_framespecified, the pedestrian tracker will publish poses in the tf frame of the camera:*_depth_optical_frame