A multi-scale UAV aerial image object detection model MS-YOLOv7 based on YOLOv7 was proposed to address the issues of a large number of objects and a high proportion of small objects that commonly exist in the Unmanned Aerial Vehicle (UAV) aerial image. The new network is developed with a multiple detection head and a CBAM convolutional attention module to extract features at different scales. To solve the problem of high-density object detection, a YOLOv7 network architecture combined with the Swin Transformer units is proposed, and a new pyramidal pooling module, SPPFS is incorporated into the network. Finally, we incorporate the SoftNMS and the Mish activation function to improve the network’s ability to identify overlapping and occlusion objects. Various experiments on the open-source dataset VisDrone2019 reveal that our new model brings a significant performance boost compared to other state-of-the-art (SOTA) models. Compared with the YOLOv7 object detection algorithm of the baseline network, the $\mathrm{mAP}0.5$ of MS-YOLOv7 increased by 6.0%, the $\mathrm{mAP}0.95$ increased by 4.9%. Ablation experiments show that the designed modules can improve detection accuracy and visually display the detection effect in different scenarios. This experiment demonstrates the applicability of the MS-YOLOv7 for UAV aerial photograph object detection.

With the rapid development of science and technology, uncrewed aerial vehicle (UAV) technology has shown a wide range of application prospects in various fields. The accuracy and real-time performance of UAV target detection play a vital role in ensuring safety and improving the work efficiency of UAVs. Aimed at the challenges faced by the current UAV detection field, this paper proposes the Gathering Cascaded Dilated DETR (GCD-DETR) model, which aims to improve the accuracy and efficiency of UAV target detection. The main innovations of this paper are as follows: (1) The Dilated Re-param Block is creatively applied to the dilatation-wise Residual module, which uses the large kernel convolution and the parallel small kernel convolution together and fuses the feature maps generated by multi-scale perception, greatly improving the feature extraction ability, thereby improving the accuracy of UAV detection. (2) The Gather-and-Distribute mechanism is introduced to effectively enhance the ability of multi-scale feature fusion so that the model can make full use of the feature information extracted from the backbone network and further improve the detection performance. (3) The Cascaded Group Attention mechanism is innovatively introduced, which not only saves the computational cost but also improves the diversity of attention by dividing the attention head in different ways, thus enhancing the ability of the model to process complex scenes. In order to verify the effectiveness of the proposed model, this paper conducts experiments on multiple UAV datasets of complex scenes. The experimental results show that the accuracy of the improved RT-DETR model proposed in this paper on the two UAV datasets reaches 0.956 and 0.978, respectively, which is 2% and 1.1% higher than that of the original RT-DETR model. At the same time, the FPS of the model is also improved by 10 frames per second, which achieves an effective balance between accuracy and speed.

Image denoising is a significant challenge in computer vision. While many models perform well in low‐noise environments, their denoising capabilities are relatively weak under high‐noise conditions. In addition, these models often overlook the robustness issues under adversarial attacks, leading to a marked decrease in denoising stability when facing malicious attacks. To address the challenges of achieving consistently high‐quality denoising in both high‐noise and low‐noise environments, adapting to various complex scenarios with high robustness, and enhancing the model’s resilience against attacks, we propose the NOP‐GAN, a powerful image denoising model. This model modifies the GAN architecture by integrating a U‐Net with a pyramid coordinate attention mechanism and a noise optimization algorithm into a generator of the GAN. Experimental results demonstrate that the NOP‐GAN possesses superior performance in denoising tasks and robustness against adversarial attacks.

In recent years, with the continuous development of autonomous driving technology, 3D object detection has naturally become a key focus in the research of perception systems for autonomous driving. As the most crucial component of these systems, 3D object detection has gained significant attention. Researchers increasingly favor the deep learning framework Transformer due to its powerful long-term modeling ability and excellent feature fusion advantages. A large number of excellent Transformer-based 3D object detection methods have emerged. This article divides the methods based on data sources. Firstly, we analyze different input data sources and list standard datasets and evaluation metrics. Secondly, we introduce methods based on different input data and summarize the performance of some methods on different datasets. Finally, we summarize the limitations of current research, discuss future directions and provide some innovative perspectives.

Image denoising has been a knotty issue in the computer vision field, although the developing deep learning technology has brought remarkable improvements in image denoising. Denoising networks based on deep learning technology still face some problems, such as in their accuracy and robustness. This paper constructs a robust denoising network based on a generative adversarial network (GAN). Since the neural network has the phenomena of gradient dispersion and feature disappearance, the global residual is added to the autoencoder in the generator network, to extract and learn the features of the input image, so as to ensure the stability of the network. On this basis, we proposed an optimization algorithm (OA), to train and optimize the mean and variance of noise on each node of the generator. Then the robustness of the denoising network was improved through back propagation. Experimental results showed that the model’s denoising effect is remarkable. The accuracy of the proposed model was over 99% in the MNIST data set and over 90% in the CIFAR10 data set. The peak signal to noise ratio (PSNR) and structural similarity (SSIM) values of the proposed model were better than the state-of-the-art models in the BDS500 data set. Moreover, an anti-interference test of the model showed that the defense capacities of both the fast gradient sign method (FGSM) and project gradient descent (PGD) attacks were significantly improved, with PSNR and SSIM values decreased by less than 2%.