AVDNet: A Small-Sized Vehicle Detection Network for Aerial Visual Data

TL;DR

AVDNet is a novel one-stage network designed to improve small vehicle detection in aerial images by using residual blocks and a new dataset, achieving superior accuracy and efficiency over existing methods.

Contribution

The paper introduces AVDNet with residual blocks and a new aerial dataset, enhancing small vehicle detection accuracy and robustness in aerial imagery.

Findings

AVDNet outperforms state-of-the-art methods in mAP.

The network demonstrates robustness across multiple aerial datasets.

The new ABD dataset provides valuable data for future research.

Abstract

Detection of small-sized targets in aerial views is a challenging task due to the smallness of vehicle size, complex background, and monotonic object appearances. In this letter, we propose a one-stage vehicle detection network (AVDNet) to robustly detect small-sized vehicles in aerial scenes. In AVDNet, we introduced ConvRes residual blocks at multiple scales to alleviate the problem of vanishing features for smaller objects caused because of the inclusion of deeper convolutional layers. These residual blocks, along with enlarged output feature map, ensure the robust representation of the salient features for small sized objects. Furthermore, we proposed a recurrent-feature aware visualization (RFAV) technique to analyze the network behavior. We also created a new airborne image data set (ABD) by annotating 1396 new objects in 79 aerial images for our experiments. The effectiveness of AVDNet is validated on VEDAI, DLR- 3K, DOTA, and the combined (VEDAI, DLR-3K, DOTA, and ABD) data set. Experimental results demonstrate the significant performance improvement of the proposed method over state-of-the-art detection techniques in terms of mAP, computation, and space complexity.