GithubHelp home page GithubHelp logo

yukunxia / superpoint-stereo-visual-odometry Goto Github PK

View Code? Open in Web Editor NEW
24.0 2.0 2.0 188.43 MB

Deployment and evaluation code for a SuperPoint-based Stereo Visual Odometry

License: MIT License

CMake 12.04% C++ 85.67% Python 2.29%
superpoint stereo-vision visual-odometry tensorrt jetson kitti neural-network

superpoint-stereo-visual-odometry's Introduction

SuperPoint Stereo Visual Odometry

This repository stores the evaluation and deployment code of our On-Device Machine Learning course project. Team members are Yukun Xia, and Yuqing Qin. The goal of this project is to explore the relationships between the performance and latency/energy efficiency of different visual odometry front ends, and benchmark them on a Jetson Nano. We retrained SuperPoint models with different backbones, as well as changed the configurations, eg. input resolution and quantization. We tested 78 configurations on KITTI dataset, and our temporary conclusion is that, for an edge device with a basic DL accelerator, the performance, latency and energy efficiency of a SuperPoint model can be competitive against classic methods, eg. ORB and SIFT. But still, we need more data and more rigorous tests.

For more details, please read our project report pdf under the root folder.

Dependencies

  • ROS
  • OpenCV 4.5.4
  • Ceres
  • Eigen3 3.3
  • CUDA
  • TensorRT 8.2

System Diagram

Usage

Basic

  • Download rosbags or create rosbags from KITTI dataset via kitti2bag
  • catkin_make
  • source devel/setup.bash or source devel/setup.zsh
  • rosrun odml_visual_odometry engine_generation.py
  • roslaunch odml_data_processing main.launch
  • rostopic pub /kitti_loader_action_server/goal odml_data_processing/kitti_data_loaderActionGoal "header: seq: 0 stamp: secs: 0 nsecs: 0 frame_id: '' goal_id: stamp: secs: 0 nsecs: 0 id: '' goal: kitti_eval_id: 4 description: ''"

Evaluate over all configurations

please checkout the long_term_eval branch

superpoint-stereo-visual-odometry's People

Contributors

yukunxia avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar

superpoint-stereo-visual-odometry's Issues

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.