Current deep learning‐based facial expression recognition mainly focused on the six basic human emotions and relied on large‐scale and well‐annotated data. For complex emotion recognition, such a large amount of data are not easy to obtain, and a high‐quality annotation is even more difficult. Therefore, in this paper, we regard complex emotion recognition via facial expressions as a few‐shot learning problem and introduce a metric‐based few‐shot model named self‐cure relation networks (SCRNet), which is robust to label noises and is able to classify facial images of new classes of emotions by only few examples from each. Specifically, SCRNet learns a distance metric based on deep features abstracted by convolutional neural networks and predicts a query image’s emotion category by computing relation scores between the query image and the few examples of each new class. To tackle the label noise problem, SCRNet gives corrected labels to noise data via class prototype stored in external memory during the meta‐training phase. Experimenting on public datasets as well as on synthetic noise datasets demonstrates the effectiveness of our method.

The targets in the aerial view of UAVs are small, scenes are complex, and background noise is strong. Additionally, the low computational capability of UAVs is challenged when trying to meet the requirements of large neural networks. Therefore, a lightweight object detection algorithm tailored for UAV platforms, called RSG-YOLO, is proposed. The algorithm introduces an attention module constructed with receptive field attention and coordinate attention, which helps reduce background noise interference while improving long-range information dependency. It also introduces and refines a fine-grained downsampling structure to minimize the loss of target information during the downsampling process. A general efficient layer aggregation network enhances the base feature extraction module, improving gradient flow information. Additionally, a detection layer rich in small target information is added, while redundant large object detection layers are removed, achieving a lightweight design while enhancing detection accuracy. Experimental results show that, compared to the baseline algorithm, the improved algorithm increases the P, R, mAP@0.5, and mAP@0.5:0.95 by 6.9%, 7.2%, 8.4%, 5.8%, respectively, on the VisDrone 2019 dataset, and by 5.7%, 9%, 9.3%, 3.6%, respectively, on the TinyPerson dataset, while reducing the number of parameters by 23.3%. This significantly enhances the model’s detection performance and robustness, making it highly suitable for object detection tasks on low-computing-power UAV platforms.

Topology control technology is one of the key technologies in wireless sensor network (WSN). To solve the problem of excessive energy consumption of single-hop communication between cluster head and coordinator in WSN after clustering, a cluster head multi-hop topology control algorithm based on geographical location is presented in this paper. This algorithm first divides the whole region into a number of radial fan regions centered on the base station. Each cluster head judges its own fan region by its geographical location and searches the base station in the same sector area to find its upper hop cluster head to form a multi-hop backbone network to communicate with the base station. Thus, the energy consumption pressure of cluster heads far away from base stations can be alleviated and the network lifetime can be prolonged. The simulation results show that the algorithm has obvious advantages in balancing energy consumption of network nodes and prolonging network lifetime.

A multi-scale segmentation algorithm based on high resolution satellite remote sensing image is proposed in this study. Firstly, scale parameter was set up according to the size of the target object. Then the heterogeneity between adjacent image regions was calculated by the formula. Thereafter, the heterogeneity and scale parameter were compared. If heterogeneity was smaller than the scale parameter, the adjacent image regions were merged and the process of comparing and merging continued; otherwise; if heterogeneity was not smaller than the scale parameter, the process of merging the adjacent image regions ended and image segmentation of remote sensing image completed. Finally, the target objects were divided into single image object. Our experiment shows that this novel algorithm can adjust scale parameter according to the size of the target object, and achieve multi scale segmentation of remote sensing image.