DOTA-DOAI

Abstract

This repo is the codebase for our team to participate in DOTA related competitions, including rotation and horizontal detection. We mainly use FPN-based two-stage detector, and it is completed by Xue Yang and Jirui Yang.

We also recommend two rotated/oriented object detection benchmarks, which are led by Xue Yang and Yue Zhou.

Performance

DOTA1.0 (Task1)

Model	Backbone	Training data	Val data	mAP	Model Link	Tricks	lr schd	Data Augmentation	GPU	Image/GPU	Configs
FPN	ResNet152_v1d (600,800,1024)->MS	DOTA1.0 trainval	DOTA1.0 test	78.99	model	ALL	2x	Yes	2X GeForce RTX 2080 Ti	1	cfgs_dota1.0_res152_v1.py

DOTA1.0 (Task2)

Model	Backbone	Training data	Val data	mAP	Model Link	Tricks	lr schd	Data Augmentation	GPU	Image/GPU	Configs
FPN (memory consumption)	ResNet152_v1d (600,800,1024)->MS	DOTA1.0 trainval	DOTA1.0 test	81.23	model	ALL	2x	Yes	2X Quadro RTX 8000	1	cfgs_dota1.0_res152_v1.py

Visualization

Performance of published papers on DOTA datasets

DOTA1.0 (Task1)

Model	Backbone	mAP	Paper Link	Code Link	Remark	Recommend
FR-O (DOTA)	ResNet101	52.93	CVPR2018	MXNet	DOTA dataset, baseline	✅
IENet	ResNet101	57.14	arXiv:1912.00969	-	anchor free
TOSO	ResNet101	57.52	ICASSP2020	-	geometric transformation
Wang et al.	HRNet	58.29	Comput. Electr. Eng.	-	anchor free
PIoU Loss	DLA-34	60.5	ECCV2020	PyTorch	IoU loss, anchor free	✅
R²CNN	ResNet101	60.67	arXiv:1706.09579	TF	scene text, multi-task, different pooled sizes, baseline	✅
RRPN	ResNet101	61.01	TMM arXiv:1703.01086	TF	scene text, rotation proposals, baseline	✅
Axis Learning	ResNet101	65.98	Remote Sensing	-	single stage, anchor free	✅
Li et al.	ResNet50	66.01	IGARSS2021	-	refine, feature alignment
MARNet	ResNet101	67.19	IJRS	-	based on scrdet
ICN	ResNet101	68.16	ACCV2018	-	image cascade, multi-scale	✅
GSDet	ResNet101	68.28	TIP	-	scale reasoning
RADet	ResNeXt101	69.09	Remote Sensing	-	enhanced FPN, mask rcnn
KARNET	ResNet50	68.87	CISNRC 2020	-	attention denoising, anchor refining
RoI Transformer	ResNet101	69.56	CVPR2019	MXNet, PyTorch, MMRotate	roi transformer	✅
CAD-Net	ResNet101	69.90	TGRS arXiv:1903.00857	-	attention
ProbIoU	ResNet50	70.04	arXiv:2106.06072	TF	gaussian bounding boxes, hellinger distance	✅
ROTP	ResNet101	70.29	arXiv:2202.06565	-	solar corona heatmap, key points, head of object
A²S-Det	ResNet101	70.64	Remote Sensing	-	label assign
AOOD	ResNet101	71.18	Neural Computing and Applications	-	attention + R-DFPN
CGP Box	ResNet18	71.35	IJRS	-	center-guide points
Zhou et al.	CSPDarknet53	71.5	ijgi	-	enhanced fpn
ACE	DLA34	71.7	TIP	-	corner point	✅
Pei et al.	ResNet101	71.76	IGRASS2021	-	enhanced FPN
Cascade-FF	ResNet152	71.80	ICME2020	-	refined retinanet + feature fusion
SCPNet	Hourglass104	72.20	GRSL	-	corner points
P-RSDet	ResNet101	72.30	Access	-	anchor free, polar coordinates	✅
Zhang et al.	ResNet101	72.37	GSIS	-	refine-stage
ROPDet	ResNet101-DCN	72.42	J REAL-TIME IMAGE PR	-	point set representation
SCRDet	ResNet101	72.61	ICCV2019	TF: R²CNN++, IoU-Smooth L1: RetinaNet-based, R³Det-based	attention, angular boundary problem	✅
O²-DNet	Hourglass104	72.8	ISPRS, arXiv:1912.10694	-	centernet, anchor free	✅
HRPNet	HRNet-W48	72.83	GRSL	-	polar
SARD	ResNet101	72.95	Access	-	IoU-based weighted loss
GLS-Net	ResNet101	72.96	Remote Sensing	-	attention, saliency pyramid
ProjBB	ResNet101	73.03	Access	code, codebase	new definition of bounding box
DRN	Hourglass104	73.23	CVPR	code	centernet, feature selection module, dynamic refinement head, new dataset (SKU110K-R)	✅
FADet	ResNet101	73.28	ICIP2019	-	attention
RBA-CenterNet	ResNet101	73.41	IJCNN	-	centernet, refine feature
MFIAR-Net	ResNet152	73.49	Sensors	-	feature attention, enhanced FPN
CFC-NET	ResNet101	73.50	TGRS	PyTorch	critical feature, label assign, refine	✅
Dual-Det	ResNet18	73.62	IJRS	code	keypoint-based
Li et al.	CSP-Hourglass	73.70	GRSL	-	CSP-Hourglass Net
R³Det	ResNet101	73.79	AAAI2021	TF, r3det-on-mmdetection, r3det-pytorch, MMRotate	refined single stage, feature alignment	✅
SDCDet	ResNet101	73.89	PRAI2021	-	instance segmentation direction correction
SegmRDet	ResNet50	74.14	Neurocomputing	-	segmentation-baed, new training and inference
Hou et al.	ResNet101	74.44	TIP	-	enhanced FPN, feature alignment
Wu et al.	ResNet50	74.45	J Electron Imaging	-	enhanced FPN, feature alignment
CenterRot	ResNet152	74.75	Remote Sensing	-	anchor free, deformable-fpn, csl
MEAD	ResNet101	74.80	Applied Intelligence	-	mechanism anchor free, mask guided, refine feature
TS⁴Net	ResNet101	74.82	Neurocomputing	-	label assign
FEDet	ResNet50	74.89	ICCSE 2021	-	refine feature, angle constraint
SRep-RDet	RepVGG-B1g2	74.89	IJRS	-	refine feature, attention, repvgg
Yuan et al.	HRNet32	74.97	ICCTIT2021	-	centernet-r
Gliding Vertex	ResNet101	75.02	TPAMI arXiv:1911.09358	PyTorch, MMRotate	quadrilateral bbox	✅
OSSDet	ResNeXt101	75.08	JSTARS	-	feature enhancement and alignment
EFN	U-Net	75.27	Preprints	-	Field-based	✅
SAR	ResNet152	75.26	Access	-	boundary problem	✅
TricubeNet	Hourglass104	75.26	WACV2022	code	2D tricube kernel	✅
Mask OBB	ResNeXt101	75.33	Remote Sensing	-	attention, multi-task	✅
BBAVectors	ResNet101	75.36	WACV2021	PyTorch	keypoint based	✅
SAOA	ResNet101	75.41	ICIG2021	-	anchor free, spatial self-attention
Zand et al.	DarkNet53	75.5	TGRS	-	angle classification
MSFF	-	75.60	ICCECE 2022	-	multi-scale feature fusion
RODFormer	ViT-B4	75.60	Sensors	-	vision transformer
FFA	ResNet101	75.7	ISPRS	-	enhanced FPN, rotation proposals
CBDA-Net	DLA-34-DCN	75.74	TGRS	-	dual attention
APE	ResNeXt101(32x4)	75.75	TGRS arXiv:1906.09447	-	adaptive period embedding, length independent IoU (LIIoU)	✅
R⁴Det	ResNet152	75.54	Image Vis Comput	-	feature recursion and refinement
SurroundNet	ResNet152	75.88	Remote Sensing	-	attention, anchor free, quad
RIE	HRGANet-W48	75.94	Remote Sensing	-	center-based rotated inscribed ellipse
F³-Net	ResNet152	76.02	Remote Sensing	-	feature fusion and filtration
CenterMap OBB	ResNet101	76.03	TGRS	-	center-probability-map
DA-Net	ResNet101	76.11	GRSL	-	feature alignment
MDL-p	ResNet101	76.16	arXiv:2204.00840	-	gaussian modeling
CSL	ResNet152	76.17	ECCV2020	TF, MMRotate, Pytorch: YOLOv5_DOTA_OBB (CSL)	angular boundary problem	✅
MRDet	ResNet101	76.24	TGRS	-	arbitrary-oriented rpn, multiple subtasks
AFC-Net	ResNet101	76.27	Neurocomputing	-	adaptive feature concatenate
RSDet/RSDet++	ResNet152	76.30	AAAI2021/TCSVT	TF	quadrilateral bbox, angular boundary problem	✅
OWSR	Ensemble	76.36	CVPR2019 WorkShop	-	enhanced FPN
SLA	ResNet50	76.36	Remote Sensing	PyTorch	sparse label assignment	✅
OPLD	ResNet101	76.43	J-STARS	PyTorch	boundary problem, point-guided	✅
Polar Ray	ResNet101	76.50	ACM MM2021	-	polar rays representation	✅
SIoU	ResNet50	76.54	JSTARS	-	splicing intersection over union	✅
R³Det++	ResNet152	76.56	arXiv:2004.13316	TF	refined single stage, feature alignment, denoising	✅
PolarDet	ResNet101	76.64	IJRS arXiv:2010.08720	-	polar, center-semantic	✅
Beyond Bounding-Box	ResNet152	76.67	CVPR2021	PyTorch, MMRotate	point-based, reppoints	✅
SCRDet++	ResNet101	76.81	TPAMI	TF	angular boundary problem, denoising	✅
DAFNe	ResNet101	76.95	arXiv:2109.06148	PyTorch	single stage, anchor free, center-to-corner regression
DAL+S²A-Net	ResNet50	76.95	AAAI2021	PyTorch	label assign	✅
GGHL	DarkNet53	76.95	TIP	PyTorch	gaussian heatmap labeling	✅
EAutoDet	DarkNet53	77.05	arXiv:2203.10747	-	nas, yolov5	✅
Yu et al.	Res2Net50	77.18	JSTARS	-	boundary-aware vectors, centernet
DCL	ResNet152	77.37	CVPR2021	TF	boundary problem	✅
MSFF	ResNet50	77.46	JSTARS	-	rotation invariance features
RIDet	ResNet50	77.62	GRSL	PyTorch, TF	quad., representation ambiguity	✅
Oriented RepPoints	Swin-Tiny	77.63	CVPR2022	PyTorch	point-based, reppoints	✅
RSP	ViTAEv2-S	77.72	arXiv:2204.02825	PyTorch	remote sensing pretrain	✅
RDD	ResNet101	77.75	Remote Sensing	PyTorch	rotation-decoupled
CenterOBB	DLA-34	77.85	Remote Sensing	-	angle classification, centernet	✅
FSDet	ResNet50	77.85	TGRS	-	label assign, feature refinemen, anchor free	✅
CG-Net	ResNet101	77.89	arXiv:2103.11399	PyTorch	attention
HSP	ResNet101	78.01	TGRS	-	hierarchical semantic propagation
FoRDet	VGG16	78.13	TGRS	-	refinenet
AProNet	ResNet101	78.16	ISPRS	Pyrotch	axis projection-based angle learning, feature enhancement
TransConvNet	Swin Transformer	78.41	Remote Sensing	-	enhanced fpn, self attention, transformer
FR-Est	ResNet101-DCN	78.49	TGRS	-	point-based estimator	✅
DARDet	ResNet50	78.74	GRSL	PyTorch	varifocalnet, dcn, piou
FCOSR	ResNeXt101	78.90	arXiv:2111.10780	PyTorch	anchor free
P2P	ResNet101	79.15	AAAI2022	-	poly iou loss	✅
SASM	ResNeXt101	79.17	AAAI2022	MMRotate, PyTorch	label assign	✅
AO2-DETR	ResNet50	79.22	arXiv:2205.12785	PyTorch	detr, feature refinement	✅
GSNet	ResNet101	79.37	arXiv:2204.02033	-	enhanced fpn
S²A-Net	ResNet50/ResNet101	79.42/79.15	TGRS	PyTorch, MMRotate	refined single stage, feature alignment	✅
OFA-Net	ResNet101	79.52	PRICAI2021	-	refined single stage, feature alignment
TARDet	ResNet101	79.57	CVPRW2022	-	anchor free, feature alignment
RBox	ResNet50	79.59	CVPR2022	-	transformer, feature sampling and grouping	✅
O²DETR	ResNet50	79.66	arXiv:2106.03146	-	deformable detr, transformer	✅
ROSD	ResNet101	79.76	Access	-	refined single stage, feature alignment
RAOD	ResNet101	79.78	Applied Intelligence	-	enhanced fpn, feature refine
SES-Net	ResNet50	79.80	arXiv:2111.03420	-	sampling equivariance, self-attention	✅
SARA	ResNet50/ResNet101	79.91/79.13	Remote Sensing	-	self-adaptive aspect ratio anchor, refine
ARP+R-EIoU	YOLOv5x6	79.93	arXiv:2109.10187	-	area ratio of parallelogram, R-EIoU, yolov5
GF-CSL	ResNet101	79.94	TGRS	PyTorch	circular smooth label, gaussian focal loss	✅
ADT-Det	ResNet152	79.95	Remote Sensing	-	feature pyramid transformer, feature refinement
ReDet	ReR50-ReFPN	80.10	CVPR2021	PyTorch, MMRotate	rotation-equivariant, rotation-invariant roI align	✅
G-Rep	Swin-Tiny	80.12	arXiv:2205.11796	MMRotate	pointset-based, gaussian modeling	✅
GWD	ResNet152	80.23	ICML2021	TF, MMRotate	boundary discontinuity, square-like problem, gaussian wasserstein distance loss	✅
DEA	ReR50-ReFPN	80.37	TGRS	PyTorch	dynamic enhancement anchor	✅
O2MER	ResNet50	80.43	arXiv:2112.00504	-	consistent geometric constraint
DODet	ResNet50	80.62	TGRS	PyTorch	oriented proposal network, localization-guided head
KLD	ResNet152	80.63	NeurIPS2021	TF, MMRotate	Kullback-Leibler divergence, high-precision, scale invariance	✅
AOPG	ResNet50/ResNet101	80.66/80.19	arXiv:2110.01931	PyTorch	anchor free, feature align
Li et al.	ResNet101	80.68	Remote Sensing	-	enhance fpn, polar
AFF-Det	ResNet50	80.73	Acm T Multim Comput.	-	enhanced fpn
Oriented R-CNN	ResNet50/ResNet101	80.87/80.52	ICCV2021	PyTorch, MMRotate	Rotation FPN, Gliding Vertex
OSKDet	ResNet101	80.91	CVPR2022	-	keypoint localization	✅
KFIoU	Swin-Tiny	80.93	arXiv:2201.12558	TF, MMRotate	Gaussian modeling, kalman filter	✅
Point RCNN	ReR50-ReFPN/Swin-Tiny	80.71/81.32	Remote Sensing	-	point based, cascade rcnn	✅

DOTA1.0 (Task2)

Model	Backbone	mAP	Paper Link	Code Link	Remark	Recommend
FR-H (DOTA)	ResNet101	60.46	CVPR2018	MXNet	DOTA dataset, baseline	✅
Deep Active Learning	ResNet18	64.26	arXiv:2003.08793	-	CenterNet, Deep Active Learning	✅
SBL	ResNet50	64.77	arXiv:1810.08103	-	single stage
CenterFPANet	ResNet18	65.29	HPCCT & BDAI 2020 arXiv:2009.03063	-	light-weight
MARNet	ResNet101	71.73	IJRS	-	based on scrdet
FMSSD	VGG16	72.43	TGRS	-	IoU-based weighted loss, enhanced FPN
ICN	ResNet101	72.45	ACCV2018	-	image cascade, multi-scale	✅
IoU-Adaptive R-CNN	ResNet101	72.72	Remote Sensing	-	IoU-based weighted loss, cascade
EFR	VGG16	73.49	Remote Sensing	PyTorch	enhanced FPN
AF-EMS	ResNet101	73.97	Remote Sensing	-	scale-aware feature, anchor free
SCRDet	ResNet101	75.35	ICCV2019	TF	attention, angular boundary problem	✅
FADet	ResNet101	75.38	ICIP2019	-	attention
MFIAR-Net	ResNet152	76.07	Sensors	-	feature attention, enhanced FPN
F³-Net	ResNet152	76.48	Remote Sensing	-	feature fusion and filtration
Mask OBB	ResNeXt101	76.98	Remote Sensing	-	attention, multi-task	✅
CenterMap OBB	ResNet101	77.33	TGRS	-	center-probability-map
ASSD	VGG16	77.8	TGRS	-	feature aligned
AFC-Net	ResNet101	78.06	Neurocomputing	-	adaptive feature concatenate
CG-Net	ResNet101	78.26	arXiv:2103.11399	PyTorch	attention
OPLD	ResNet101	78.35	J-STARS	PyTorch	boundary problem, point-guided	✅
A²RMNet	ResNet101	78.45	Remote Sensing	-	attention, enhanced FPN, different pooled sizes
OWSR	Ensemble (ResNet101 + ResNeXt101 + mdcn-ResNet101)	78.79	CVPR2019 WorkShop	-	enhanced FPN
NPMMR-Det	DarkNet53	78.93	TGRS	PyTorch	non-local, enhanced FPN
Parallel Cascade R-CNN	ResNeXt101	78.96	Journal of Physics: Conference Series	-	cascade rcnn
DM-FPN	ResNet-Based	79.27	Remote Sensing	-	enhanced FPN
DDBN	ResNet101	79.3	PR	-	adjacent feature compensation, dual detection branch
SCRDet++	ResNet101	79.35	TPAMI	TF	denoising	✅
HSP	ResNet101	80.42	TGRS	-	hierarchical semantic propagation
AFF-Det	ResNet101	81.18	Acm T Multim Comput.	-	enhanced fpn

DOTA1.5 (Task1)

Model	Backbone	mAP	Paper Link	Code Link	Remark	Recommend
DAFNe	ResNet101	71.99	arXiv:2109.06148	PyTorch	r-fcos
RAOD	ResNet101	74.70	Applied Intelligence	-	enhanced fpn, feature refine
OWSR	Ensemble (Re sNet101 + ResNeXt101 + mdcn-ResNet101)	76.60	CVPR2019 WorkShop	-	enhanced FPN
ReDet	ReR50-ReFPN	76.80	CVPR2021	PyTorch	rotation-equivariant, rotation-invariant RoI Align,	✅
APE	ResNeXt101(32x4)	78.34	TGRS arXiv:1906.09447	-	length independent IoU (LIIoU)	✅
Point RCNN	ReR50-ReFPN/Swin-Tiny	79.31/80.14	Remote Sensing	-	point based, cascade rcnn	✅

DOTA1.5 (Task2)

Model	Backbone	mAP	Paper Link	Code Link	Remark	Recommend
CDD-Net	ResNet101	61.3	GRSL	-	attention
GCWNet	ResNet101	62.7	GRSL	-	enhance fpn, feature refine
ReDet	ReR50-ReFPN	78.08	CVPR2021	PyTorch	rotation-equivariant, rotation-invariant RoI Align,	✅
OWSR	Ensemble (ResNet101 + ResNeXt101 + mdcn-ResNet101)	79.50	CVPR2019 WorkShop	-	enhanced FPN

DOTA1.5 (Task1)

Model	Backbone	mAP	Paper Link	Code Link	Remark	Recommend
Wen et al.	-	58.22	ICCECE 2022	-	yolov5, csl

Model	Paper Link	Code Link	Remark	Recommend
SSSDET	ICIP2019 arXiv:1909.00292	-	vehicle detection, lightweight
AVDNet	GRSL arXiv:1907.07477	-	vehicle detection, small object
ClusDet	ICCV2019	Caffe2	object cluster regions	✅
DMNet	CVPR2020 WorkShop	-	object cluster regions	✅
AdaZoom	arXiv:2106.10409	-	object cluster regions, reinforcement learning	✅
OIS	arXiv:1911.07732	related Pytorch code	Oriented Instance Segmentation	✅
ISOP	IGARSS2020	-	Oriented Instance Segmentation
LR-RCNN	arXiv:2005.14264	-	vehicle detection	-
GRS-Det	TGRS	-	ship detection, rotation fcos	-
DRBox	arXiv:1711.09405	Caffe	sar object detection	✅
DRBox-v2	TGRS	TF	sar object detection	-
RAPiD	arXiv:2005.11623	PyTorch	overhead fisheye images	-
OcSaFPN	arXiv:2012.09859	-	denoising	-
CR2A-Net	TGRS	-	ship detection	-
-	TGRS	-	knowledge distillation	✅
CHPDet	arXiv:2101.11189	-	new ship dataset	✅

Other Rotation Detection Codes and Benchmarks

Base Method	Code Link
RetinaNet	RetinaNet_Tensorflow_Rotation
YOLOv3	rotate-yolov3-Pytorch, YOLOv3-quadrangle-Pytorch, yolov3-polygon-Pytorch
YOLOv4	rotate-yolov4-Pytorch
YOLOv5	rotation-yolov5-Pytorch, YOLOv5_DOTA_OBB (CSL)
YOLOX	YOLOX-Pytorch
CenterNet	R-CenterNet-Pytorch
JDet	Jittor
OBBDet_Swin	OBBDet_Swin
OrientedRepPoints	OrientedRepPoints_DOTA_Swin
Localization Distillation	Rotated-LD

Dataset

Some remote sensing related object detection dataset statistics are in DATASET.md

sunlizhuang/DOTA-DOAI

DOTA-DOAI

Abstract

Performance

DOTA1.0 (Task1)

DOTA1.0 (Task2)

Visualization

Performance of published papers on DOTA datasets

DOTA1.0 (Task1)

DOTA1.0 (Task2)

DOTA1.5 (Task1)

DOTA1.5 (Task2)

DOTA1.5 (Task1)

Related Articles

Other Rotation Detection Codes and Benchmarks

Dataset