Remote Sensing Small Object Detection Network Based on Attention Mechanism and Multi-Scale Feature Fusion

Abstract

In remote sensing images, small objects have too few discriminative features, are easily confused with background information, and are difficult to locate, leading to a degradation in detection accuracy when using general object detection networks for aerial images. To solve the above problems, we propose a remote sensing small object detection network based on the attention mechanism and multi-scale feature fusion, and name it AMMFN. Firstly, a detection head enhancement module (DHEM) was designed to strengthen the characterization of small object features through a combination of multi-scale feature fusion and attention mechanisms. Secondly, an attention mechanism based channel cascade (AMCC) module was designed to reduce the redundant information in the feature layer and protect small objects from information loss during feature fusion. Then, the Normalized Wasserstein Distance (NWD) was introduced and combined with Generalized Intersection over Union (GIoU) as the location regression loss function to improve the optimization weight of the model for small objects and the accuracy of the regression boxes. Finally, an object detection layer was added to improve the object feature extraction ability at different scales. Experimental results from the Unmanned Aerial Vehicles (UAV) dataset VisDrone2021 and the homemade dataset show that the AMMFN improves the APs values by 2.4% and 3.2%, respectively, compared with YOLOv5s, which represents an effective improvement in the detection accuracy of small objects.

MFEFNet: A Multi-Scale Feature Information Extraction and Fusion Network for Multi-Scale Object Detection in UAV Aerial Images

“…In view of the various problems that have arisen, many researchers have put forward different research methods. Qu et al [24] proposed a detection head enhancement module (DHEM) that utilizes an attention mechanism and multi-scale feature fusion to enhance the representational information of objects. However, it only focuses on small objects and lacks attention to medium and large objects in UAV images.…”

Research on None-Line-of-Sight/Line-of-Sight Identification Method Based on Convolutional Neural Network-Channel Attention Module

“…Attention mechanisms have been widely used in natural language processing, data prediction, hydroacoustic identification, image segmentation, etc. Compared to deep learning network architectures, the attention mechanism is a lightweight module that tunes the network parameters by generating and assigning weights and trains the network to focus on key information to improve accuracy [24,25]. In this paper, the CAM is added to the CNN, and its structure is shown in Figure 9.…”

Improved Detection Method for Micro-Targets in Remote Sensing Images

“…The STC-YOLO algorithm performs well for traffic-sign identification. Qu et al's approach [24] introduced a feature fusion strategy based on an attention mechanism. By merging target information from different scales, this strategy enhances the semantic expression of shallow features, consequently improving the tiny-object identification capacity of the algorithm.…”

“…The algorithm proposed in Ref. [24] exhibits competent detection performance; however, its detection time is also significantly increased. PDWT-YOLO [25] is primarily designed for detecting small targets and exhibits a fast detection speed; however, it is not suitable for detecting much smaller targets, such as the targets in AI-TOD [26].…”