Deep learning has achieved great success in many fields, such as image classification and target detection. Adding small disturbance which is hard to be detected by the human eyes to original images can make the neural network output error results with high confidence. An image after adding small disturbance is an adversarial example. The existence of adversarial examples brings a huge security problem to deep learning. In order to effectively defend against adversarial examples attacks, an adversarial example defense method based on image reconstruction is proposed by analyzing the existing adversarial examples attack methods and defense methods. Our data set is based on ImageNet 1k data set, and some filtering and expansion are carried out. Four attack modes, FGSM, BIM, DeepFool and C&W are selected to test the defense method. Based on the EDSR network, multi-scale feature fusion module and subspace attention module are added. By capturing the global correlation information of the image, the disturbance can be removed, while the image texture details can be better preserved, and the defense performance can be improved. The experimental results show that the proposed method has good defense effect.
With the rapid development of deep learning, UAV target detection technology based on computer vision and artificial intelligence has been widely used in practice. However, due to the instability of UAV movement, limited by load and endurance, the development of UAV target detection is slow, and there are challenges such as significant changes in target scale, occlusion between objects, and changes in target density. This paper builds on the network model structure of YOLOv5 to address these challenges. It adds a detection head generated from low-level feature layers and high-resolution combined feature maps to detect tiny objects. We utilize the Bifpn network structure and a weighted fusion splicing approach to fuse more features and introduce an improved Coordinate Attention to obtain location information for feature enhancement accurately. Extensive experiments on the Visdrone2021 dataset show that the model achieves good results in UAV target detection and is helpful for tiny and occluded target detection.
Convolutional neural network models have become one of the most commonly used methods for analyzing medical images. Among them, the codec structure has brought important breakthrough results for medical image segmentation. However, the current medical image segmentation method based on the codec network architecture still has many problems. The corresponding feature map of the codec network in the skip connection structure has a large semantic ambiguity, which may increase the difficulty of learning the network and reduce the segmentation performance. The codec network architecture cannot make full use of the relationship between objects in the global view, and also ignores the global context information of different scales. In this article, we add attention gate mechanism (AGs) to the jump connection structure, and introduce attention mechanism and multi-scale mechanism to solve the above problems. Our model obtains better segmentation performance while introducing fewer parameters.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.