Page
%P
-
Chapter
Correction to: Artificial Intelligence in Vision-Based Structural Health Monitoring
-
Chapter
Structural Vision Data Collection and Dataset
Images and videos are the two most commonly used data types in vision-based . The image represents the instantaneous state in the structure and the video provides continuous changes of the state of the struct...
-
Chapter
Multi-task Learning
The previous chapters have demonstrated the feasibility and adaptability of AI technologies to recognize building and infrastructure damage via images. However, many previous studies solely work on the existen...
-
Chapter
Structural Damage Localization
In addition to the classical classification tasks, structural damage localization (following the usual detection task) plays a crucial role in vision-based . As introduced in Chap. 2, damage locations are rep...
-
Chapter
Generative Adversarial Network for Structural Image Data Augmentation
As mentioned in Sect. 1.3.2 and 2.4, the practical usage of AI in SHM encounters challenges related to the scarcity of labeled data or even difficulties in data collection due to complex environments.
-
Book
-
Chapter
Vision Tasks in Structural Images
There has been an increasing trend of applying models in structural damage classification, especially in concrete crack detection and steel corrosion detection. Soukup and Huber-Mörk [1] applied to detect ste...
-
Chapter
Basics of Deep Learning
As a subset of ML, DL is attracting rapidly increasing interests nowadays. However, the study of DL dates back to the 1940s and it has undergone multiple periods of research and development [1]. The concept of DL...
-
Chapter
Active Learning
Manually labeling data is time-consuming. Therefore, it is usually difficult to obtain a large amount of labeled data compared to having unlabeled data. If the SSL method is employed, both labeled and unlabele...
-
Chapter
Structural Image Classification
As mentioned in Chap. 2, structural image classification is one of the most fundamental tasks in vision-based . However, the introduction of technologies into the field of is not straightforward. Therefor...
-
Chapter
Interpreting CNN in Structural Vision Tasks
In previous chapters, promising results have been achieved using AI, e.g., DL-based models, in vision-based SHM problems. However, the internal working principle in AI, especially for the DL model, is hard to ...
-
Chapter
Structural Damage Segmentation
Structural damage segmentation is another key task in vision-based SHM. As introduced in Chap. 2, images are labeled pixel by pixel and segmentation algorithms and models aim to recognize all pixels, group a r...
-
Chapter
Introduction
(SHM), as defined in Lynch and Loh (The shock and vibration digest (2006), [1]), is a “paradigm” of rapidly identifying structural damage in an instrumented structural system and can be classified in terms of th...
-
Chapter
Basics of Machine Learning
In the past decades, has achieved significant advances in both theory and applications. Generally speaking, has three major categories: supervised learning, unsupervised learning, and reinforcement learning b...
-
Chapter
Semi-Supervised Learning
Previous chapters demonstrate the effectiveness of ML and DL under the supervised learning setting, where all training data are well-labeled, refer to Sect. 3.1.
-
Article
Response Spectrum Code-Conforming PEER PBEE using Stochastic Dynamic Analysis and Information Theory
In this paper, the tools of the stochastic dynamic analysis are adopted for Performance-Based Earthquake Engineering (PBEE). The seismic excitation is defined through a evolutionary Power Spectral Density comp...
-
Article
Shaking table test method of building curtain walls using floor capacity demand diagrams
Seismic demands of a curtain wall (CW) relate to the floor response of the main structure. These demands include both acceleration and deformation. Floor capacity spectrum combines these demands into one accel...
-
Article
Open AccessThe 3rd Global Summit of Research Institutes for Disaster Risk Reduction: Expanding the Platform for Bridging Science and Policy Making
The Global Alliance of Disaster Research Institutes held its 3rd Global Summit of Research Institutes for Disaster Risk Reduction at the Disaster Prevention Research Institute, Kyoto University, Japan, 19–21 ...
-
Chapter
A Decision Support Tool for Sustainable and Resilient Building Design
In this chapter, an integrated approach for a holistic (involving notions of safety, resiliency and sustainability) building design is presented to select the optimal design alternative based on multiple confl...
-
Chapter
Progressive Collapse Simulation of Vulnerable Reinforced Concrete Buildings
There are many vulnerable reinforced concrete (RC) buildings located in earthquake-prone areas around the world. These buildings are characterized by the lack of seismic details and corresponding non-ductile b...
