Sign inSign up

loaywael/se_ssd

By loaywael

•Updated about 5 years ago

SE-SSD is a 3d-object-detection model: https://github.com/Vegeta2020/SE-SSD

Image
1

97

loaywael/se_ssd repository overview

⁠Usage

creating a container from this image.

    $ docker run -itd --name SE-SSD --gpus all --ipc host -v <host-shared-path>:/shared_area loaywael/se_ssd:ros-kinetic-ub16
    $ docker exec -it SE-SSD bash

make sure of setting the ros-network between the container and the host

    $ ifconfig
    $ nano ~/.bashrc
    $ export ROS_MASTER_URI=http://<host-ip>:11311
    $ export ROS_IP=<container-ip>
    $ roslaunch se_ssd detect_3d_objects.launch    # default kitti-bag for arl-bag set arl:=true

run a demo script

    $ cd SE-SSD/tools/
    $ python <script-name>

in case of using rosbag/rviz run them on the host.


⁠The image supports:
  • model inference given a directory of PointCloud binaries.
  • integrated with ROS
  • supports custom datasets
⁠To Do:
  • support multi-class prediction
  • support 360 degree detection

Tag summary

Content type

Image

Digest

Size

5.7 GB

Last updated

about 5 years ago

docker pull loaywael/se_ssd:ros-kinetic-ub16