This paper mainly discusses an intelligent analysis stage based on a large number of computerscience and Internet. This intelligent analysis stage based on a large number of Internet includes: an intelligent statisti...
详细信息
In order to improve the efficiency of indoor mobile robots in locating and segmenting environmental instances, an instance segmentation method based on RTMDet is proposed. Firstly, the more powerful ConvNeXt V2 is use...
详细信息
Steel, being a widely utilized material in industrial production, holds a pivotal role in ensuring product safety and longevity. Hence, the exploration and implementation of steel surface defect detection technology c...
详细信息
Multimodal data often requires manual annotation for training, and the annotation process is tedious and time-consuming. The lack of labels makes it difficult to build large-scale labeled multi-modal datasets. In resp...
详细信息
Due to the complexity and time-varying nature of the environment, single image recognition technology is difficult to guarantee the high-precision perception of the target when affected by ground object occlusion, ill...
详细信息
ISBN:
(纸本)9798350386783;9798350386776
Due to the complexity and time-varying nature of the environment, single image recognition technology is difficult to guarantee the high-precision perception of the target when affected by ground object occlusion, illumination environment, observation angle and so on, and cannot make full use of the correlation of captured data between multiple perceptions. To solve this problem, this paper proposes a multi-view data fusion method based on an evaluation network. By using the complementary of data between multiple perceptions, the target can still be accurately perceived when some perceptions cannot provide effective information. The algorithm in this paper has achieved good recognition results in the public datasets, and has achieved effective recognition of targets 100m away in the datasets constructed by ourselves.
Due to the problems of redundant features and high dimensionality in collaborative intrusion detection, the direct use of the Sand Cat Swarm Optimization (SCSO) algorithm performs poorly and is prone to fall into loca...
详细信息
ISBN:
(纸本)9798350349184;9798350349191
Due to the problems of redundant features and high dimensionality in collaborative intrusion detection, the direct use of the Sand Cat Swarm Optimization (SCSO) algorithm performs poorly and is prone to fall into local optimality. Thus, this paper proposes a Collaborative-Enhanced Sand Cat Optimization (CESCSO) algorithm to solve these problems for the purpose of promoting collaborative intrusion detection. Firstly, a better balance between the two phases of exploration and exploitation is achieved by introducing a non-linear transformation factor, which lays the foundation for the global search for the optimal solution. Secondly, the golden sine strategy is utilized to prevent individuals from falling into local optimality and improve the exploration ability of the SCSO algorithm. Thirdly, escaping behaviors are added to improve exploration efficiency and accuracy. Finally, a random opposition-based learning strategy is introduced at the end of each iteration to mutate the population and select the global optimal solution. The proposed CESCSO algorithm is then tested and compared with the original SCSO algorithm and other classical algorithms in six benchmark functions. The results show that the proposed CESCSO algorithm is highly competitive in terms of convergence speed and accuracy, and can significantly reduce the number of data dimensions and effectively improve the network intrusion detection performance.
This article proposes a multilevel relational distillation method, which focuses more on the spatial structure relationship of feature maps compared to traditional detection distillation. Compared with advanced networ...
详细信息
Breast microcalcifications are one of the important indicators of early breast cancer. Radiologists discriminate benign and malignant microcalcifications mainly based on their morphology and distribution. Convolutiona...
详细信息
Portable sinks assume a vital part in information assortment within wireless sensor networks (WSNs), offering dynamic and flexible solutions to address the limitations of static sink-based architectures. Unlike their ...
详细信息
In response to the stealth and difficulty in detecting the characteristics of Advanced Persistent Threat (APT) attacks, as well as the blind spots in internal network supervision, this paper proposes a multidimensiona...
详细信息
暂无评论