Home

Awesome

RATE: Real-time Asynchronous Feature Tracking with Event Cameras

This repo is for event-based continuous corner detection and tracking described in the following paper:

Mikihiro Ikura, Cedric Le Gentil, Marcus G. Müller, Florian Schuler, Atsushi Yamashita and Wolfgang Stürzl: "RATE: Real-time Asynchronous Feature Tracking with Event Cameras", Proceedings of the 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS2024), Abu Dhabi (UAE), October 2024. Paper

https://github.com/user-attachments/assets/6a164201-4383-4537-b3a5-40d049043c34

ros packages lists

Requirements

Before starting to run

Update submodule

git submodule update --init --recursive

Build docker image

docker build --build-arg UID=$(id -u) --build-arg GID=$(id -g) -t rate:latest .

Run docker container

docker run -it --privileged --net=host --rm -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix:rw -v /path_to_rosbag_data_in_host/:/app/rosbag --name rate rate:latest

or

docker compose up -d
docker exec -it rate /bin/bash

Build catkin_ws (in docker terminal)

cd ~/catkin_ws
catkin build --cmake-args -DCMAKE_BUILD_TYPE=Release
source devel/setup.bash

Open X server (host in new terminal)

xhost local:docker

How to run RATE, continuous feature detection and tracking with event rosbag data

1st terminal (in docker container)

roslaunch haste_ros haste_fd_timeslice_sae.launch event_topic:=/dvs/events camera_size:=240x180 camera_calib:=/home/rate/catkin_ws/src/haste_ros/haste/dataset/calib.txt best_tracker:=alignment_score

2nd terminal (in host if you have ros environment locally / otherwise in docker container)

rosbag play boxes_6dof.bag