10000 GitHub - collector-m/Fade: Fade3D : Fast and Deployable 3D Object Detection for Autonomous Driving (TITS2025)
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content
forked from wayyeah/Fade

Fade3D : Fast and Deployable 3D Object Detection for Autonomous Driving (TITS2025)

License

Notifications You must be signed in to change notification settings

collector-m/Fade

10000
 
 

Repository files navigation

Fade 3D: Fast and Deployable 3D Object Detection for Autonomous Driving(TITS 2025)

Visualization of real traffic scenario testing

detect Video: https://github.com/wayyeah/Fade/blob/master/detect.mp4

KITTI 3D Object Detection Baseline

< 733F td align="center">51.5
FPS (RTX3090) FPS (Jetson Orin) Car_R40@0.7 Easy Car_R40@0.7 Mod. Car_R40@0.7 Hard download download(TensorRT)
Fade 12.4 90.92 82.00 77.49 model-50M model-25M

Getting Started

Please refer to GETTING_STARTED.md to learn more usage about this project.

TensorRT How to use

A. prepare environment

  1. install CUDA>=11.4
  2. install TensorRT>8.5.x.x and modified TensorRT Path in fade_trt/CMakeLists.txt
  3. modified path in fade_trt/main.cpp Line 137 model_path and Line 138 path

B. compile and run

  1. cd fade_trt
  2. mkdir build & cd build
  3. cmake .. & make
  4. ./demo kitti

We deployed our Fade object detection algorithm on a car equipped with an OS-128 LiDAR and Jetson Orin, and achieved autonomous driving of the car.

Video: https://github.com/wayyeah/Fade/blob/master/AutonomousDriving.mp4

About

Fade3D : Fast and Deployable 3D Object Detection for Autonomous Driving (TITS2025)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 85.9%
  • Cuda 7.3%
  • C++ 6.3%
  • Other 0.5%
0