AI Paper Atlas

Research map / Benchmarks / DOTA

DOTA leaderboard

12 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on DOTA. Best mAP: 79.155 (RetinaNet-P2P Loss#, AAAI 2022).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

DOTA · mAP

12 papers, each with its best reported number for its own method; higher is better.

#mAP ↑MethodPaperTopic
179.155RetinaNet-P2P Loss#
RetinaNet ResNet101-DCN
Polygon-to-Polygon Distance Loss for Rotated Object Detection
AAAI 2022 · text, Comparison of State-of-t
CNN features
277.63Oriented RepPoints
Swin-T-FPN
Oriented RepPoints for Aerial Object Detection
CVPR 2022 · DOTA results table
CNN features
376.95S^2A-Net+DAL
S^2A-Net, R-50
Dynamic Anchor Learning for Arbitrary-Oriented Object Detection
AAAI 2021 · Table: Performance evaluation
CNN features
476.67CFA
ResNet-152
Beyond Bounding-Box: Convex-Hull Feature Adaptation for Oriented and Densely Packed Object Detection
CVPR 2021 · Table 1
CNN features
576.47R³Det
ResNet152
R3Det: Refined Single-Stage Detector with Feature Refinement for Rotating Object
AAAI 2021 · Table: Detection accuracy on D
CNN features
676.28Oriented R-CNN
R-101-FPN
Oriented R-CNN for Object Detection
ICCV 2021 · Table 2
CNN features
774.92RepPoints + BCL + SA-S + SA-M
RepPoints + BCL + SA-S
Shape-Adaptive Selection and Measurement for Oriented Object Detection
AAAI 2022 · Table 2
CNN features
874.1RSDet+ResNet152+Refine
RetinaNet-H
Learning Modulated Loss for Rotated Object Detection
AAAI 2021 · DOTA results table
CNN features
969.56RoI Transformer with FPN
ResNet101
Learning RoI Transformer for Oriented Object Detection in Aerial Images
CVPR 2019 · Table 3
CNN features
1058.6CastDet
Faster R-CNN RN50-C4
Toward Open Vocabulary Aerial Object Detection with CLIP-Activated Student-Teacher Learning
ECCV 2024 · Table 5(b)
Open-set and few-shot detection
2 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.