3d_harmonic_loss_for_object_detection

3D Harmonic Loss: Towards Task-consistent and Time-friendly 3D Object Detection on Edge for V2X Orchestration

https://github.com/xjtu-haolin/3d_harmonic_loss_for_object_detection

Science Score: 54.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
    Found CITATION.cff file
  • codemeta.json file
    Found codemeta.json file
  • .zenodo.json file
    Found .zenodo.json file
  • DOI references
  • Academic publication links
    Links to: ieee.org
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (7.9%) to scientific vocabulary
Last synced: 6 months ago · JSON representation ·

Repository

3D Harmonic Loss: Towards Task-consistent and Time-friendly 3D Object Detection on Edge for V2X Orchestration

Basic Info
  • Host: GitHub
  • Owner: XJTU-Haolin
  • License: mit
  • Language: Python
  • Default Branch: main
  • Homepage:
  • Size: 12.7 MB
Statistics
  • Stars: 9
  • Watchers: 1
  • Forks: 2
  • Open Issues: 0
  • Releases: 0
Created over 3 years ago · Last pushed over 1 year ago
Metadata Files
Readme License Citation

README.md

3D Harmonic Loss: Towards Task-consistent and Time-friendly 3D Object Detection on Edge for V2X Orchestration

Notification

Our proposed 3D Harmonic Loss can be applied to many lidar-based 3D object detection methods for solving inconsistency problem as shown below.

Alleviating inconsistency problem in 3D detection via our proposed 3D harmonic loss

Paper in IEEE: PDF (IEEE T-VT)

Our implementation is relied on mmdetection3D

Environment

python = 3.7
pytorch = 1.6
CPU: i7-10700K
GPU: RTX-2080Ti
other requriements are same as in mmdetection3D

Dataset Preparation

Please follow the mmdetection3D to convert KITTI Dataset and Waymo Dataset

Training (MMdetection3D)

Note that our 3D harmonic loss optimization can be implemented to train almost all anchor-based 3D detectors without inference time cost. Please get familar with mmdetection3D-format config in advance, and then you can check configs/_base_/models/hv_second_secfpn_kitti_harmonic_loss.py as an example to customize other model configs, and use our revised anchor-head with harmonic loss for 3D detectors. You can follow official document of mmdetection3D to know how to configurate configs and train models.

Training (OpenPCDet)

will be released soon...

Test

You can follow official document of mmdetection3D to test models.

TODO Lists

  • [X] Readme Completion
  • [X] Paper Preprinted
  • [X] Support KITTI Dataset
  • [X] Support Waymo Dataset
  • [ ] Support OpenPCDet benchmark

Owner

  • Name: Haolin Zhang
  • Login: XJTU-Haolin
  • Kind: user

Citation (CITATION.cff)

cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
  - name: "MMDetection3D Contributors"
title: "OpenMMLab's Next-generation Platform for General 3D Object Detection"
date-released: 2020-07-23
url: "https://github.com/open-mmlab/mmdetection3d"
license: Apache-2.0

GitHub Events

Total
Last Year

Dependencies

docker/Dockerfile docker
  • pytorch/pytorch ${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel build
docker/serve/Dockerfile docker
  • pytorch/pytorch ${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel build
requirements/build.txt pypi
requirements/docs.txt pypi
  • docutils ==0.16.0
  • m2r *
  • mistune ==0.8.4
  • myst-parser *
  • sphinx ==4.0.2
  • sphinx-copybutton *
  • sphinx_markdown_tables *
requirements/mminstall.txt pypi
  • mmcv-full >=1.4.8,<=1.6.0
  • mmdet >=2.24.0,<=3.0.0
  • mmsegmentation >=0.20.0,<=1.0.0
requirements/optional.txt pypi
  • open3d *
  • spconv *
  • waymo-open-dataset-tf-2-1-0 ==1.2.0
requirements/readthedocs.txt pypi
  • mmcv >=1.4.8
  • mmdet >=2.24.0
  • mmsegmentation >=0.20.1
  • torch *
  • torchvision *
requirements/runtime.txt pypi
  • lyft_dataset_sdk *
  • networkx >=2.2,<2.3
  • numba ==0.53.0
  • numpy *
  • nuscenes-devkit *
  • plyfile *
  • scikit-image *
  • tensorboard *
  • trimesh >=2.35.39,<2.35.40
requirements/tests.txt pypi
  • asynctest * test
  • codecov * test
  • flake8 * test
  • interrogate * test
  • isort * test
  • kwarray * test
  • pytest * test
  • pytest-cov * test
  • pytest-runner * test
  • ubelt * test
  • xdoctest >=0.10.0 test
  • yapf * test
requirements.txt pypi
setup.py pypi