Zhaoyu Su

• Doctor of Philosophy
• Hong Kong University of Science and Technology

Two research studies have been implemented to explore the potential of applying artificial intelligence (AI) technologies in works projects and maintenance work of the Drainage Services Department (DSD) for enhancing the efficiency related to environmental monitoring and structural inspection, referred to as the AIEIA and AIBIM projects, respective...

This paper presents a framework for automated defect inspection of the concrete structures, made up of data collection, defect detection, scene reconstruction, defect assessment and data integration stages. A mobile data collection system, comprising a 360° camera and a digital Light Detection and Ranging (LiDAR), is developed to render high flexib...

In this work, we propose a novel two-stage framework for the efficient 3D point cloud object detection. Instead of transforming point clouds into 2D bird eye view projections, we parse the raw point cloud data directly in the 3D space yet achieve impressive efficiency and accuracy. To achieve this goal, we propose dynamic voxelization, a method tha...

This paper reports the use of the deep learning-based technique to characterize the particle orientation of clay samples. The U-Net model was applied to perform semantic segmentation for identifying individual kaolinite particles, based on the scanning electron microscopic images taken from clay samples subjected to 1-D consolidation. The measurabl...

Drainage network extraction is essential for different research and applications. However, traditional methods have low efficiency, low accuracy for flat regions, and difficulties in detecting channel heads. Although deep learning techniques have been used to solve these problems, different challenges remain unsolved. Therefore, we introduced distr...

Automatic bird detection in ornithological analyses is limited by the accuracy of existing models, due to the lack of training data and the difficulties in extracting the fine-grained features required to distinguish bird species. Here we apply the domain randomization strategy to enhance the accuracy of the deep learning models in bird detection....

Inspection of civil infrastructure is a major challenge to engineers due to the limitations in existing practice, which are as laborious, time-consuming and prone to error. To address these issues, we have applied deep learning for image-based inspection of concrete defects of civil infrastructure, and have established an artificial intelligence-em...

This paper reports a feasible alternative to compile a landslide inventory map (LIM) from remote sensing datasets using the application of an artificial intelligence–driven methodology. A deep convolutional neural network model, called LanDCNN, was developed to generate segmentation maps of landslides, and its performance was compared with the benc...

3D point cloud interpretation is a challenging task due to the randomness and sparsity of the component points. Many of the recently proposed methods like PointNet and PointCNN have been focusing on learning shape descriptions from point coordinates as point-wise input features, which usually involves complicated network architectures. In this work...

This paper reports the application of deep learning for implementing the anomaly detection of defects on concrete structures, so as to facilitate the visual inspection of civil infrastructure. A convolutional autoencoder was trained as a reconstruction-based model, with the defect-free images, to rapidly and reliably detect defects from the large v...

Discriminator from generative adversarial nets (GAN) has been used by some research as feature extractor in transfer learning and worked well. But there are also some studies believed that this is a wrong research direction because intuitively the task of discriminator focuses on separating the real samples from the generated ones, making the featu...

In this study, a novel method for pore network extraction in the weakly consolidated media is reported, which unifies both the Delaunay tessellation (DT) and maximal ball (MB) methods in a complementary way. This unified method retains the advantages of both methods, and most importantly, eliminates the disadvantages when either the DT or the MB me...