Skip to content

Deep learning-based diver detection and tracking modules. #RAL2018 #ICRA2019

Notifications You must be signed in to change notification settings

xahidbuffon/deep-diver-following

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

59 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This is a container for an autonomous diver-following project. Deep object detection models are used for diver (and other objects such as ROV) detection. A simplified version of that is utilized for autonomous tracking (and following) of a single diver by an underwater robot. The ROS version, tested on Aqua-8 robot, is provided in the diver_following_cnn folder.

Testing the detector

Use the test_detector.py file for testing detection performances on individual images.

Single diver Multiple divers Divers and ROVs
det-7 det-7 det-1

Testing the diver-tracker

For testing diver-tracking on a video or sequences of images, use the test_diver_tracker.py file. A couple of videos and image sequences are provided in the test_data folder. Change the argument values to test other files.

frame t frame t+n frame t+2n
det-86 det-96 det-106
det-1118 det-1138 det-1158

ROS version

@article{islam2018towards,
  title={{Toward a Generic Diver-Following Algorithm: Balancing Robustness and Efficiency in Deep Visual Detection}},
  author={Islam, Md Jahidul and Fulton, Michael and Sattar, Junaed},
  journal={{IEEE Robotics and Automation Letters (RA-L)}},
  volume = {4},
  number = {1},
  pages = {113--120},
  year={2018},
  publisher={IEEE}
}

Releases

No releases published

Packages

No packages published