DOTA: A Large-Scale Dataset for Object Detection in Aerial Images
By Gui-Song Xia, Xiang Bai, Jian Ding, Zhen Zhu, Serge J. Belongie, Jiebo Luo, Mihai Datcu, Marcello Pelillo, Liangpei Zhang. 2017
2,943 citations recorded by the source corpus.
Object detection is an important and challenging problem in computer vision. Although the past decade has witnessed major advances in object detection in natural scenes, such successes have been slow to aerial imagery, not only because of the huge variation in the scale, orientation and shape of the object instances on the earth's surface, but also due to the scarcity of well-annotated datasets of objects in aerial scenes. To advance object detection research in Earth Vision, also known as Earth Observation and Remote Sensing, we introduce a large-scale Dataset for Object deTection in Aerial images (DOTA). To this end, we collect 2806 aerial images from different sensors and platforms. Each image is of the size about 4000 × 4000 pixels and contains objects exhibiting a wide variety of scales, orientations, and shapes. These DOTA images are then annotated by experts in aerial image interpretation using 15 common object categories. The fully annotated DOTA images contains 188, 282 instances, each of which is labeled by an arbitrary (8 d.o.f.) quadrilateral. To build a baseline for object detection in Earth Vision, we evaluate state-of-the-art object detection algorithms on DOTA. Experiments demonstrate that DOTA well represents real Earth Vision applications and are quite challenging.
Subfields: Computer Vision and Pattern Recognition, Media Technology, Evaluation & Benchmarks, Remote Sensing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Ming-Hsuan Yang
- Xiaogang Wang
- Yi Yang
- Trevor Darrell
- Chunhua Shen
- Ling Shao
In Media Technology
In Evaluation & Benchmarks
- Guisong Xia
- Vladimir Risojevic
- Vladan Stojnić
- Xue Yang
- Junchi Yan
- Marco Körner
- Wangmeng Zuo
- Dacheng Tao
In Remote Sensing
Citations in the corpus
- 3D Instance Segmentation of MVS Buildings (2021, background): “(DOTA) [64], few of them focus on instance segmentation at the pixel level.”
- Large Selective Kernel Network for Remote Sensing Object Detection (2023, background): “0 [61] consists of 2,806 remote sensing images.”
- RingMo: A Remote Sensing Foundation Model With Masked Image Modeling (2023, background): “1) Statistical Information: Some of the images in the dataset are selected from public datasets in the RS field, such as [4], [5], [37], [40], [64], [65], and the other images are obtained from China’s Gaofen-2 satellite.”
- Object Detection in Optical Remote Sensing Images: A Survey and A New Benchmark (2019, methodology): “9) DOTA: The DOTA (Xia et al., 2018) is a new large ‐ scale geospatial object detection dataset, which consists of 15 different object categories: baseball diamond, basketball court, bridge, harbor, helicopter, ground track field, large vehicle, plane, ship, small vehicle, soccer ball field, storage”
- Align Deep Features for Oriented Object Detection (2020, methodology): “This architecture enjoys high-computational efficiency but often lags behind in accuracy [3].”
- ReDet: A Rotation-equivariant Detector for Aerial Object Detection (2021, methodology): “With the development of general object detection, many well-designed methods have been proposed for oriented object detection [1, 7, 24, 35, 38, 40, 42], showing promising performance on challenging datasets [21, 35].”
- A Survey of Deep Learning-Based Object Detection (2019, background): “[167] G.-S. Xia, X. Bai, J. Ding, Z. Zhu, S. Belongie, J. Luo, M. Datcu, M. Pelillo, and L. Zhang, ‘‘DOTA: A large-scale dataset for object detection in aerial images,’’ in Proc.”
- Rethinking Rotated Object Detection with Gaussian Wasserstein Distance Loss (2021, methodology): “The rotation detector R 3 Net [67] achieves the state-of-the-art performance on large-scale DOTA.”
- Land-cover classification with high-resolution remote sensing images using transferable deep models (2018, background): “scene classification [25, 29], object detection [30], image retrieval [31–33], as well as land-cover classification [34–39].”
- TTST: A Top-k Token Selective Transformer for Remote Sensing Image Super-Resolution (2024, neutral): “0 [64], DIOR [65] and NWPU-RESISC45 [66].”
- A Comprehensive Review of YOLO Architectures in Computer Vision: From YOLOv1 to YOLOv8 and YOLO-NAS (2023, background): “Adding an extra prediction head, feature fusion layers, and a residual transformer module, NRT-YOLO improved YOLOv5l by 5.4% in the DOTA data set [134].”
- R3Det: Refined Single-Stage Detector with Feature Refinement for Rotating Object (2019, background): “The benchmark DOTA [34] ferent sensors and platforms.”
- Towards Multi-class Object Detection in Unconstrained Remote Sensing Imagery (2018, methodology): “f results on NWUH VHR-10 and UCAS-AOD datasets. method train data test data mAP Cheng et al.[34] NWUH VHR-10 NWUH VHR-10 72.63 ours NWUH VHR-10 NWUH VHR-10 95.01 ours DOTA NWUH VHR-10 82.23 Xia et al.[4] UCAS-AOD UCAS-AOD 89.41 ours UCAS-AOD UCAS-AOD 95.67 ours DOTA UCAS-AOD 86.13 4 Conclusions In t”
- Self-Supervised Learning of Remote Sensing Scene Representations Using Contrastive Multiview Coding (2021, methodology): “Besides that, using NWPU-DOTA dataset for self-supervised pre-training can give results that are better than the results obtained using representations learned by supervised learning on ImageNet.”
- Object Detection and Image Segmentation with Deep Learning on Earth Observation Data: A Review-Part I: Evolution and Recent Trends (2020, methodology): “The adjacent RoI align does extract much more object specific features from this rotated RoIs resulting in state of the art performance on the DOTA dataset [138].”
- Object Detection in 20 Years: A Survey (2019, background): “an/NWPUVHR10dataset.html 204 LEVIR [91] 2018 Consists of ˘22,000 Google Earth images and ˘10,000 independently labeled targets (airplane, ship, oil-pot). url: https://pan.baidu.com/s/1geTwAVD 15 DOTA [92] 2018 The first remote sensing detection dataset to incorporate rotated bounding boxes. Consists ”