mirajull / OrientedRepPoints

The code for “Oriented RepPoints for Aerial Object Detection (CVPR 2022)”

Oriented RepPoints for Aerial Object Detection

Installation

Either refer to for installation and dataset preparation or build a conda environment from orp_env.yml

Getting Started

Please see for the basic usage.

Results and Models

The results on DOTA test set are shown in the table below. More detailed results please see the paper.

Model	Backbone	data aug(HSV+Rotation)	mAP	model	log
OrientedReppoints	R-50		75.97	model	log
OrientedReppoints	R-101		76.52	model	log
OrientedReppoints	Swin-Tiny	√	78.11	model	log

Few Shot Learning with OrientedRepPoints on Omniglot Data

1. Place the data folder according to the notebook data folder setting or configure the data path
2. Run the jupyter notebook FSL_OrientedRepPoints_Omniglot.ipynb

Few Shot Learning with OrientedRepPoints on our Custom Energy Infrastructure Data

1. Place the data folder according to the notebook data folder setting or configure the data path
2. Run the jupyter notebook FSL_OrientedRepPoints_Custom.ipynb

Acknowledgements

Here are some great resources we benefit. We would espeicially thank the authors of:

MMdetection

RepPoints

AerialDetection

BeyondBoundingBox