14 papers with code • 1 benchmarks • 4 datasets
Object Detection in Aerial Images is the task of detecting objects from aerial images.
( Image credit: DOTA: A Large-Scale Dataset for Object Detection in Aerial Images )
In this paper, we present a large-scale Dataset of Object deTection in Aerial images (DOTA) and comprehensive baselines for ODAI.
Object detection in aerial images is an active yet challenging task in computer vision because of the bird's-eye view perspective, the highly complex backgrounds, and the variant appearances of objects.
Considering the shortcoming of feature misalignment in existing refined single-stage detector, we design a feature refinement module to improve detection performance by getting more accurate features.
Specifically, a sampling fusion network is devised which fuses multi-layer feature with effective anchor sampling, to improve the sensitivity to small objects.
The fully annotated DOTA images contains $188, 282$ instances, each of which is labeled by an arbitrary (8 d. o. f.)
To address this issue, in this work we extend the horizontal keypoint-based object detector to the oriented object detection task.
Ranked #1 on Object Detection on DOTA
However most of existing methods rely on heuristically defined anchors with different scales, angles and aspect ratios and usually suffer from severe misalignment between anchor boxes and axis-aligned convolutional features, which leads to the common inconsistency between the classification score and localization accuracy.
More precisely, we incorporate rotation-equivariant networks into the detector to extract rotation-equivariant features, which can accurately predict the orientation and lead to a huge reduction of model size.
With the newly introduced DAL, we achieve superior detection performance for arbitrary-oriented objects with only a few horizontal preset anchors.
Ranked #1 on 2D Object Detection on DOTA (using extra training data)
Especially when detecting densely packed objects in aerial images, methods relying on horizontal proposals for common object detection often introduce mismatches between the Region of Interests (RoIs) and objects.