Alexey Merzlyakov

Thanks for the idea! Honestly, since all 3 tracking approaches are tied to current and previous frame, just changing the camera position might be unstable. So, re-localization at key-frame position...

Added work-in-progress PR to `openvslam` part for setting the camera to a known pose as recommended in https://github.com/OpenVSLAM-Community/openvslam_ros/issues/11#issuecomment-814032968. The development for `openvslam_ros` part is in progress and the second PR...

Duplicating the Slack, just for the case: The SGF smoother was estimated with SmacHybrid-A* and default NavFn path planner. On both, it produces pretty aligned smoothed path, except of its...

I've verified the SG path smoother with NavFn, ThetaStar and Smac-HybridA* (DUBIN). Now all cases are working pretty well, smoothing the local factured paths produced by path planners. Here are...

In Collision Monitor, we have already subscribing on footprints in case of APPROACH behavior model. So, I do not see any obstacle to have an ability to subscribe to any...

Good. The major thing here was agreed: The current functionality of polygons in the Collision Monitor to be enhanced by subscribing to the topics: the similar way as `APPROACH` model...

Hi, @tonynajjar. From the perspective of use-cases that it would be a nice functionality expansion, to have general information about shapes triggered that will be back to user; and action...

Hi! In its current state, Collision Monitor has the `Polygon` and `Source` (and all their derived classes) to be operate independently from CM node logic itself. They have the following...

@fujitatomoya, thank you for the making an attention on this issue! Yes, it will handle the immediate problem, so I would like to see these two PR-s to be merged....

Sure. For simple testing I did not use any objects detecting / ML front-end. Instead of this, I've just treated all laser scan points, each as obstacles. So, number of...