Multi-Modality Fusion &amp; Inductive Knowledge Transfer Underlying Non-Sparse Multi-Kernel Learning and Distribution Adaption

With the massive construction of bridge infrastructures, bridge health monitoring systems have gradually matured in application and research, but previous research has primarily focused on structural damage detection and bridge safety warnings based on valid data. The structural details of steel bridge panels and structural systems are determined by the coupling effects of many intrinsic and extrinsic uncertainties, such as material properties, structural characteristics, manufacturing processes, and random traffic loads. The evaluation of fatigue is a difficult task. This article first builds a big data platform, utilizing its high-efficiency parallel computing capability and highly fault-tolerant distributed file system to achieve second-level monitoring data processing; ensuring real-time data cleaning, data analysis, and safety warning; and building a big data analysis and processing platform with high reliability, high availability, high storage efficiency, and high scalability of bridge health monitoring. The big data platform chooses HDFS for offline data storage and Spark for data analysis and modelling after comparing and analysing the benefits and drawbacks of various big data technologies. Kafka is used for caching real-time data, and Spark-streaming is used for reading data and real-time processing. Finally, the platform’s superiority and reliability in terms of offline computing performance, real-time online performance, scalability, and fault tolerance are confirmed through experimental analysis; the optimal data cleaning method is derived by comparing and analysing monitoring data noise, jump point, and drift phenomena. This part of the research is based on bridge temperature data with stable signals and bridge strain data with fluctuating signals, taking into account the influence of different data types; the corresponding data missing repair algorithms are proposed for different types of data to form a complete and general data patching method process. The probabilistic fracture mechanics theory, in comparison to the traditional deterministic fatigue assessment method, can better reflect the essential uncertainty of fatigue problems and is an effective way to assess the fatigue performance of orthotropic steel bridge decks. The goal of data patching is to ensure data recovery accuracy of over 90%, with no patching repair required for monitoring data with too much missing data. The endurance life of bridge structures is predicted using a big data probabilistic statistics approach based on a variety of factors such as material properties, construction characteristics, manufacturing processes, and random traffic loads.

Section: A Probabilistic Statistics-based Approach To Bridge Life Pre...mentioning

confidence: 99%

Probabilistic Statistics-Based Endurance Life Prediction of Bridge Structures

Zhang

2022

“…In recent years, Few-Shot learning is a subfield of machine learning [ 7 , 8 ]. The definition of machine learning is that a computer program learns from experience (experiences, E) to solve a certain task (task, T) and perform a certain performance measure (performance, P).…”

Section: Few-shot Learning Based On Contrastive Average Lossmentioning

confidence: 99%

Dance Fitness Action Recognition Method Based on Contour Image Spatial Frequency Domain Features and Few-Shot Learning

Chi

2022

In recent years, the research work of artificial intelligence technology has progressed rapidly, and various classic Few-Shot learning models have achieved unprecedented success in many artificial intelligence application fields. These include face recognition, object classification detection and tracking, speech recognition, and natural language processing, which greatly facilitate our lives. This paper aims to identify dance fitness movements based on contour image spatial frequency domain features and Few-Shot learning technology. This paper proposes a Few-Shot learning method based on contrastive average loss for Few-Shot learning. This method makes the learned model more representative by improving the loss function and performing a normalization process, and it proposes a feature extraction algorithm that combines improved LBP and HOG for action recognition technology. The experimental results show that the recognition accuracy of the algorithm in this paper is 93.10%, 90.30%, and 92.70% for walking, opening hands, and running, respectively. This illustrates the effectiveness of the fusion feature algorithm.

“…The following are some of its innovations: There is basically no complete set of music data expression rules in algorithmic composition system. Based on the in-depth study of related literature, this paper proposes a new algorithm composition network from the perspective of machine learning algorithm [ 5 – 7 ]. It restrains the style and quality of generated music by using music theory rules to construct a reasonable Reward function.…”

Section: Introductionmentioning

confidence: 99%

Algorithm Composition and Emotion Recognition Based on Machine Learning

2022

This paper proposes a new algorithm composition network from the perspective of machine learning, based on an in-depth study of related literature. At the same time, this paper examines the characteristics of music and develops a model for recognising musical emotions. Using the model’s information entropy of pitch and intensity to extract the main melody track, note features are extracted from bar features. Finally, the cosine of the vector included angle is used to judge the similarity between feature vectors of several adjacent sections, allowing the music to be divided into several independent segments. The emotional model of music is used to analyze each segment’s emotion. By quantifying music features, this paper classifies and quantifies music emotion based on the mapping relationship between music features and emotion. Music emotion can be accurately identified by the model. The model’s emotion recognition accuracy is up to 93.78 percent, and the algorithm’s recall rate is up to 96.3 percent, according to simulation results. The recognition method used in this paper has a higher recognition ability than other methods, and the emotion recognition result is more reliable. This paper can not only meet the composer’s auxiliary creative needs, but it can also help intelligent music services.