LR-CNN: Local-aware Region CNN for Vehicle Detection in Aerial Imagery

TL;DR

LR-CNN is a novel two-stage vehicle detection method in aerial images that improves accuracy for dense, small, and arbitrarily oriented targets by enhancing local feature invariance and addressing boundary quantization issues.

Contribution

The paper introduces LR-CNN, which enhances translation and local feature invariance for better detection of dense, small, and arbitrarily oriented vehicles in aerial imagery, overcoming limitations of existing methods.

Findings

Significant improvement over state-of-the-art on VEDAI and DOTA datasets.

Effective detection of dense, small, and arbitrarily oriented vehicles.

Good generalization demonstrated on the DLR 3K dataset.

Abstract

State-of-the-art object detection approaches such as Fast/Faster R-CNN, SSD, or YOLO have difficulties detecting dense, small targets with arbitrary orientation in large aerial images. The main reason is that using interpolation to align RoI features can result in a lack of accuracy or even loss of location information. We present the Local-aware Region Convolutional Neural Network (LR-CNN), a novel two-stage approach for vehicle detection in aerial imagery. We enhance translation invariance to detect dense vehicles and address the boundary quantization issue amongst dense vehicles by aggregating the high-precision RoIs' features. Moreover, we resample high-level semantic pooled features, making them regain location information from the features of a shallower convolutional block. This strengthens the local feature invariance for the resampled features and enables detecting vehicles in an arbitrary orientation. The local feature invariance enhances the learning ability of the focal loss function, and the focal loss further helps to focus on the hard examples. Taken together, our method better addresses the challenges of aerial imagery. We evaluate our approach on several challenging datasets (VEDAI, DOTA), demonstrating a significant improvement over state-of-the-art methods. We demonstrate the good generalization ability of our approach on the DLR 3K dataset.