Cross-view image matching aims to match images of the same target scene acquired from different platforms. With the rapid development of drone technology, cross-view matching by neural network models has been a widely accepted choice for drone position or navigation. However, existing public datasets do not include images obtained by drones at different heights, and the types of scenes are relatively homogeneous, which yields issues in assessing a model’s capability to adapt to complex and changing scenes. In this end, we present a new cross-view dataset called SUES-200 to address these issues. SUES-200 contains 24120 images acquired by the drone at four different heights and corresponding satellite view images of the same target scene. To the best of our knowledge, SUES-200 is the first public dataset that considers the differences generated in aerial photography captured by drones flying at different heights. In addition, we developed an evaluation for efficient training, testing and evaluation of cross-view matching models, under which we comprehensively analyze the performance of nine architectures. Then, we propose a robust baseline model for use with SUES-200. Experimental results show that SUES-200 can help the model to learn highly discriminative features of the height of the drone.

Visual geo-localization plays a crucial role in positioning and navigation for unmanned aerial vehicles, whose goal is to match the same geographic target from different views. This is a challenging task due to the drastic variations in different viewpoints and appearances. Previous methods have been focused on mining features inside the images. However, they underestimated the influence of external elements and the interaction of various representations. Inspired by multimodal and bilinear pooling, we proposed a pioneering feature fusion network (MBF) to address these inherent differences between drone and satellite views. We observe that UAV’s status, such as flight height, leads to changes in the size of image field of view. In addition, local parts of the target scene act a role of importance in extracting discriminative features. Therefore, we present two approaches to exploit those priors. The first module is to add status information to network by transforming them into word embeddings. Note that they concatenate with image embeddings in Transformer block to learn status-aware features. Then, global and local part feature maps from the same viewpoint are correlated and reinforced by hierarchical bilinear pooling (HBP) to improve the robustness of feature representation. By the above approaches, we achieve more discriminative deep representations facilitating the geo-localization more effectively. Our experiments on existing benchmark datasets show significant performance boosting, reaching the new state-of-the-art result. Remarkably, the recall@1 accuracy achieves 89.05% in drone localization task and 93.15% in drone navigation task in University-1652, and shows strong robustness at different flight heights in the SUES-200 dataset.

Cross-view geo-localization has been widely used as an important technique for determining the geographical location of unmanned aerial vehicles (UAVs). Despite various image retrieval methods proposed, drone and satellite image cross-view geo-localization still remain challenging due to their wildly inconsistent view angles. In this article, we propose a new framework, the Swin-radial-locality network (SRLN), to extract robust image feature representations. Specifically, SRLN is based on a pruned version of the Swin transformer, which integrates multiscale feature aggregation within a Siamese network structure, featuring shared weights and equipped with multiclassification heads. SRLN is mainly comprised of a radial-slicer-network (RSN) and a local-pattern-network (LPN), which is designed to effectively harmonize directional information from drone-captured images and broader environmental features from satellite imagery, crucial for capturing angle and feature details between drone and satellite images. The RSN part focuses on capturing fine-grained features that represent the drone’s directional information, while the LPN is utilized for a more comprehensive analysis of broader environmental features. Extensive experiments are carried out on widely used public benchmark datasets, i.e., University-1652 and SUES-200. With more than 3% improvement over existing methods in both drone-view target localization tasks and drone navigation applications, the results validate the superior performance of our multiscale feature fusion model, achieving a state-of-the-art performance record.

The past decade has demonstrated the potential of human activity recognition (HAR) with WiFi signals owing to non-invasiveness and ubiquity. Previous research has largely concentrated on enhancing precision through sophisticated models. However, the complexity of recognition tasks has been largely neglected. Thus, the performance of the HAR system is markedly diminished when tasked with increasing complexities, such as a larger classification number, the confusion of similar actions, and signal distortion To address this issue, we eliminated conventional convolutional and recurrent backbones and proposed WiTransformer, a novel tactic based on pure Transformers. Nevertheless, Transformer-like models are typically suited to large-scale datasets as pretraining models, according to the experience of the Vision Transformer. Therefore, we adopted the Body-coordinate Velocity Profile, a cross-domain WiFi signal feature derived from the channel state information, to reduce the threshold of the Transformers. Based on this, we propose two modified transformer architectures, united spatiotemporal Transformer (UST) and separated spatiotemporal Transformer (SST) to realize WiFi-based human gesture recognition models with task robustness. SST intuitively extracts spatial and temporal data features using two encoders, respectively. By contrast, UST can extract the same three-dimensional features with only a one-dimensional encoder, owing to its well-designed structure. We evaluated SST and UST on four designed task datasets (TDSs) with varying task complexities. The experimental results demonstrate that UST has achieved recognition accuracy of 86.16% on the most complex task dataset TDSs-22, outperforming the other popular backbones. Simultaneously, the accuracy decreases by at most 3.18% when the task complexity increases from TDSs-6 to TDSs-22, which is 0.14–0.2 times that of others. However, as predicted and analyzed, SST fails because of excessive lack of inductive bias and the limited scale of the training data.

With the development of autonomous driving technology, vision geo-localization has obtained a consistently growing following. How to match correct image pair from different perspectives is the key technology. Existing geo-localization methods focus on designing complex attention mechanism based on traditional backbone, e.g., VGG, ResNet, but neglect the importance of backbone network. In this article, we propose a modern backbone based geo-localization method (MBEG). MBEG introduces the latest vision fundamental network EVA-02 as backbone, which has been fully trained in large datasets. In addition, the feature rotate encoding strategy is presented to eliminate the effects of image rotation. We also apply the knowledge distillation to squeeze network's parameters for actual application. Our work exhibited excellent performance on the University-1652 dataset, and our solution attained the top-1 ranking in the UAVs in Multimedia Challenge for the University-160k dataset.

Cross-view image matching aims to match images of the same target scene acquired from different platforms. With the rapid development of drone technology, cross-view matching by neural network models has been a widely accepted choice for drone position or navigation. However, existing public datasets do not include images obtained by drones at different heights, and the types of scenes are relatively homogeneous, which yields issues in assessing a model's capability to adapt to complex and changing scenes. In this end, we present a new cross-view dataset called SUES-200 to address these issues. SUES-200 contains 24120 images acquired by the drone at four different heights and corresponding satellite view images of the same target scene. To the best of our knowledge, SUES-200 is the first public dataset that considers the differences generated in aerial photography captured by drones flying at different heights. In addition, we developed an evaluation for efficient training, testing and evaluation of cross-view matching models, under which we comprehensively analyze the performance of nine architectures. Then, we propose a robust baseline model for use with SUES-200. Experimental results show that SUES-200 can help the model to learn highly discriminative features of the height of the drone.