This paper explores the ability of Graph Neural Networks (GNNs) in learning various forms of information for link prediction, alongside a brief review of existing link prediction methods. Our analysis reveals that GNNs cannot effectively learn structural information related to the number of common neighbors between two nodes, primarily due to the nature of set-based pooling of the neighborhood aggregation scheme. Also, our extensive experiments indicate that trainable node embeddings can improve the performance of GNN-based link prediction models. Importantly, we observe that the denser the graph, the greater such the improvement. We attribute this to the characteristics of node embeddings, where the link state of each link sample could be encoded into the embeddings of nodes that are involved in the neighborhood aggregation of the two nodes in that link sample. In denser graphs, every node could have more opportunities to attend the neighborhood aggregation of other nodes and encode states of more link samples to its embedding, thus learning better node embeddings for link prediction. Lastly, we demonstrate that the insights gained from our research carry important implications in identifying the limitations of existing link prediction methods, which could guide the future development of more robust algorithms.

Deep Neural Networks (DNNs) suffer from domain shift when the test dataset follows a distribution different from the training dataset. Domain generalization aims to tackle this issue by learning a model that can generalize to unseen domains. In this paper, we propose a new approach that aims to explicitly remove domain-specific features for domain generalization. Following this approach, we propose a novel framework called Learning and Removing Domain-specific features for Generalization (LRDG) that learns a domain-invariant model by tactically removing domain-specific features from the input images. Specifically, we design a classifier to effectively learn the domain-specific features for each source domain, respectively. We then develop an encoder-decoder network to map each input image into a new image space where the learned domain-specific features are removed. With the images output by the encoder-decoder network, another classifier is designed to learn the domain-invariant features to conduct image classification. Extensive experiments demonstrate that our framework achieves superior performance compared with state-of-the-art methods.

Pipe failure prediction in the water industry aims to prioritize the pipes that are at high risk of failure for proactive maintenance. However, existing statistical or machine learning models that rely on historical failures and asset attributes can hardly leverage the structure information of pipe networks. In this work, we develop a failure prediction framework for pipe networks by jointly considering the pipes' features, the network structure, the geographical neighboring effect, and the temporal failure series. We apply a multi-hop Graph Neural Network (GNN) to failure prediction. We propose a method of constructing a geographical graph structure depending on not only the physical connections but also geographical distances between pipes. To differentiate the pipes with diverse properties, we employ an attention mechanism in the neighborhood aggregation process of each GNN layer. Also, residual connections and layer-wise aggregation are used to avoid the over-smoothing issue in deep GNNs. The historical failures exhibit a strong temporal pattern. Inspired by point process, we develop a module to learn the pipes' evolutionary effect and the time-decayed excitement of historical failures on the current state of the pipe. The proposed framework is evaluated on two real-world large-scale pipe networks. It outperforms the existing statistical, machine learning, and state-of-the-art GNN baselines. Our framework provides the water utility with core data-driven support for proactive maintenance including regular pipe inspection, pipe renewal planning, and sensor system deployment. It can be extended to other infrastructure networks in the future.

Preventing water pipe leaks and breaks has high priority for water utilities. It is a critical task for the utility to reduce water loss through leaks and breaks detection in water mains. The failure prediction and data analytics research have been conducted for an Australian water utility over the last few years to enhance the prediction of leaks and breaks detection in water mains. Intelligent sensing at sensitive locations with current research aids in prioritising investigation and prevention of potential breaks and leaks in water mains. The purpose of this work is to integrate the predictive analytics and intelligent sensing applications to identify high risk mains prior to failures. Predictive analytics and minimum night flow (MNF) analysis have been utilised to prioritise risky zones over the whole water network, and then risky pipes are identified to optimise sensors deployment. The sensing data is being collected for analysis and validation, and a machine learning model is being built based on the analysis results. This work is currently under progress and the planned outcomes will help the utility reduce water loss, improve leak detection, and enhance customer satisfaction by automating the process of leak detection using a data driven approach with smart sensors.

Monitoring drinking water quality in the entire delivery network, mainly indicated by total chlorine (TC), is a critical component of overall water supply management. However, it is extremely difficult to collect sufficient TC data from the network at customer sites, which makes it sparse for comprehensive modelling. This paper details an approach that provides TC prediction within the entire Woronora delivery network in Sydney in the next 24 hours. First, the hydraulic system is employed to capture the topology of the delivery network, so that the water travel time can be estimated using predicted water demand. The travel time links the upstream (reservoir) data to the downstream (resident) data. Then, a two-step strategy is proposed as a semi-parametric method to determine the crucial factors and build Bayesian model for TC decay to predict TC with the travel time. Lastly, the uncertainties of both data and the model are analysed to define the boundaries of prediction for better decision making. Several operational stages are involved when the approach is being deployed, including prediction interpretation, interactive tool development for water quality mapping and visualisation, and proactive optimisation. This has established a successful initiative to improve the overall water supply management for the entire Woronora delivery network.

Failure event prediction is becoming increasingly important in wide applications, such as the planning of proactive maintenance, the active investment management, and disease surveillance. To address the issue, the hazard function in survival analysis has been employed to describe the pattern of failures. Different from traditional survival analysis, this paper discovers how to apply recurrent neural network (RNN) to the long-term hazard function prediction. The proposed Long-Term RNN (LT-RNN) is able to leverage the precedent information shared by other entities, leading to more reliable long-term predictions. Specifically, our method allows a black-box treatment for modelling the hazard function which is often a pre-defined parametric form in typical survival analysis. The key idea of our approach is to model the hazard function as a nonparameteric function of the history. The same precedent information from other entities is embedded to a stitched vector for LT-RNN to automatically learn a representation of the long-term hazard function. We apply our model to the proactive maintenance problem using a large dataset from a water utility in Australia.

This paper presents a novel approach to action recognition using synthetic multi-view data from depth maps. Specifically, multiple views are first generated by rotating 3D point clouds from depth maps. A pyramid multi-view depth motion template is then adopted for multi-view action representation, characterizing the multi-scale motion and shape patterns in 3D. Empirically, despite the view-specific information, the latent information between multiple views often provides important cues for action recognition. Concentrating on this observation and motivated by the success of the dictionary learning framework, this paper proposes to explicitly learn a view-specific dictionary (called specificity) for each view, and simultaneously learn a latent dictionary (called latent correlation) across multiple views. Thus, a novel method, specificity and latent correlation learning, is put forward to learn the specificity that captures the most discriminative features of each view, and learn the latent correlation that contributes the inherent 3D information to multiple views. In this way, a compact and discriminative dictionary is constructed by specificity and latent correlation for feature representation of actions. The proposed method is evaluated on the MSR Action3D, the MSR Gesture3D, the MSR Action Pairs, and the ChaLearn multi-modal data sets, consistently achieving promising results compared with the state-of-the-art methods based on depth data.

This paper presents a vision based sign language gesture recognition framework that can assist people with impaired hearing and speech with their social interaction and interactive communications. Utilizing a low-cost sensor, such as Microsoft Kinect combined with advanced machine learning analysis, it aims to ease the challenging issue of increasing demand for professional sign language interpreting services. Specifically, this paper discusses a powerful discriminating descriptor called 3D motion map based pyramid histograms of oriented gradient (M-PHOG) which is proposed for depth-based human gesture recognition. The 3D motion map is generated through the entire depth video sequence to encode additional motion information from three projected orthogonal planes. By adding pyramid representation, HOG descriptor is extended to M-PHOG which can characterize local shapes at different spatial grid sizes for gesture recognition. The proposed approach is evaluated on MSR Gesture3D and DEVISIGN two data sets captured by depth cameras. Experimental results show that the proposed approach outperforms the current state-of-the-art methods and demonstrates the effectiveness and robustness of the proposed 3D M-PHOG descriptor. The proposed approach can translate the meaning of captured gestures as professional interpreters currently do. The novelty framework has the potential to improve the quality of life for the deaf community and reduce the communication barriers they currently experience.

This paper presents a novel framework, spatio-temporal pyramid cuboid matching (STPCM), which is designed to recognize human actions from sequences captured by depth cameras. A depth sequence is partitioned into sub-volumes and represented using pyramid motion history templates (PMHT), which maintain the multi-scale 3D motion and shape information along the temporal direction. In order to capture the spatial information of PMHT, each projected plane from PMHT is subdivided into pyramid spatio-temporal grids. We then propose a novel cuboid fusion scheme to combine spatial dependent grids from projected planes to construct pyramid cuboids that consider the 3D spatial locations in conjunction with temporal information. In the experiments, we evaluate the proposed framework on three public benchmark datasets. Experimental results demonstrate that the proposed method achieves state-of-the-art performance.

Human action recognition based on the depth maps is an important yet challenging task. In this paper, a new framework based on the 3D motion trail model (3DMTM) and Pyramid Histograms of Oriented Gradient (PHOG) is proposed to recognize human actions from sequences of depth maps. Specifically, a discriminative descriptor called 3DMTM-PHOG is proposed for depth-based human action recognition. The 3DMTM is generated through the entire depth video sequence to encode additional motion information from three projected orthogonal planes. By adding pyramid representation, Histograms of Oriented Gradient (HOG) descriptor is extended to PHOG which can well characterize local shapes at different spatial grid sizes for action recognition. PHOG is then computed from the 3DMTM as the 3DMTM-PHOG descriptor for the representation of an action. The proposed approach based on 3DMTM-PHOG descriptor is evaluated on MSR Action3D dataset captured by depth cameras. Experimental results show that the proposed approach outperforms the state-of-the-art methods and demonstrate the effectiveness and robustness of the proposed 3DMTM-PHOG descriptor.

This work presents an approach for recognizing 3D human gestures by using depth images. The proposed motion trail model (MTM) consists of both motion information and static posture information over the gesture sequence along the xoy-plane. By projecting depth images onto other two planes in 3D space, gestures can be represented with complementary information from additional planes. Accordingly 2D-MTM can be extended into 3D space in addition to the lateral scene parallel to the image plane to generate 3D-MTM. The Histogram of Oriented Gradient (HOG) is then extracted from the proposed 3D-MTM as the feature descriptor. The final recognition of gestures is performed through maximum correlation coefficient. The preliminary results demonstrate the average error rate decreases from 62.80% of baseline method to 21.74% after using the proposed approach on Chalearn gesture dataset.