Chee Peng Lim

Saeid Nahavandi, Roohallah Alizadehsani, Darius Nahavandi et al.

Automated industries lead to high quality production, lower manufacturing cost and better utilization of human resources. Robotic manipulator arms have major role in the automation process. However, for complex manipulation tasks, hard coding efficient and safe trajectories is challenging and time consuming. Machine learning methods have the potential to learn such controllers based on expert demonstrations. Despite promising advances, better approaches must be developed to improve safety, reliability, and efficiency of ML methods in both training and deployment phases. This survey aims to review cutting edge technologies and recent trends on ML methods applied to real-world manipulation tasks. After reviewing the related background on ML, the rest of the paper is devoted to ML applications in different domains such as industry, healthcare, agriculture, space, military, and search and rescue. The paper is closed with important research directions for future works.

Mehshan Ahmed Khan, Houshyar Asadi, Mohammad Reza Chalak Qazani et al.

Functional near-infrared spectroscopy (fNIRS) is employed as a non-invasive method to monitor functional brain activation by capturing changes in the concentrations of oxygenated haemoglobin (HbO) and deoxygenated haemo-globin (HbR). Various machine learning classification techniques have been utilized to distinguish cognitive states. However, conventional machine learning methods, although simpler to implement, undergo a complex pre-processing phase before network training and demonstrate reduced accuracy due to inadequate data preprocessing. Additionally, previous research in cog-nitive load assessment using fNIRS has predominantly focused on differ-sizeentiating between two levels of mental workload. These studies mainly aim to classify low and high levels of cognitive load or distinguish between easy and difficult tasks. To address these limitations associated with conven-tional methods, this paper conducts a comprehensive exploration of the im-pact of Long Short-Term Memory (LSTM) layers on the effectiveness of Convolutional Neural Networks (CNNs) within deep learning models. This is to address the issues related to spatial features overfitting and lack of tem-poral dependencies in CNN in the previous studies. By integrating LSTM layers, the model can capture temporal dependencies in the fNIRS data, al-lowing for a more comprehensive understanding of cognitive states. The primary objective is to assess how incorporating LSTM layers enhances the performance of CNNs. The experimental results presented in this paper demonstrate that the integration of LSTM layers with Convolutional layers results in an increase in the accuracy of deep learning models from 97.40% to 97.92%.

Mehshan Ahmed Khan, Houshyar Asadi, Mohammad Reza Chalak Qazani et al.

One debatable issue in traffic safety research is that cognitive load from sec-ondary tasks reduces primary task performance, such as driving. Although physiological signals have been extensively used in driving-related research to assess cognitive load, only a few studies have specifically focused on high cognitive load scenarios. Most existing studies tend to examine moderate or low levels of cognitive load In this study, we adopted an auditory version of the n-back task of three levels as a cognitively loading secondary task while driving in a driving simulator. During the simultaneous execution of driving and the n-back task, we recorded fNIRS, eye-tracking, and driving behavior data to predict cognitive load at three different levels. To the best of our knowledge, this combination of data sources has never been used before. Un-like most previous studies that utilize binary classification of cognitive load and driving in conditions without traffic, our study involved three levels of cognitive load, with drivers operating in normal traffic conditions under low visibility, specifically during nighttime and rainy weather. We proposed a hybrid neural network combining a 1D Convolutional Neural Network and a Recurrent Neural Network to predict cognitive load. Our experimental re-sults demonstrate that the proposed model, with fewer parameters, increases accuracy from 99.82% to 99.99% using physiological data, and from 87.26% to 92.02% using driving behavior data alone. This significant improvement highlights the effectiveness of our hybrid neural network in accurately pre-dicting cognitive load during driving under challenging conditions.

Yinsong Chen, Samson S. Yu, Zhong Li et al.

Post-hoc explainable AI (XAI) methods typically produce deterministic attribution maps, whereas Bayesian neural networks (BNNs) induce a distribution over explanations. Capturing the variability of this distribution is important for uncertainty-aware decision-making. This paper formalises the \emph{explanation distribution} as the push-forward measure of the BNN posterior through any Lipschitz-continuous attribution operator. It further proposes the uncertainty-aware relevance attribution operator (UA-RAO), a general family of operators that summarises the explanation distribution using the mean, variance, coefficient of variation, quantiles, and set-theoretic aggregation measures. Theoretical support is provided through Monte Carlo accessibility and Wasserstein approximation bounds. The framework is evaluated on a 15-class power quality disturbance (PQD) classification benchmark, comparing three BNN approximations paired with three attribution operators using relevance mass accuracy and intersection-over-union as localisation metrics. Results show that deep ensembles with the mean UA-RAO improve localisation over the deterministic baseline, while other UA-RAO summaries reveal uncertainty patterns absent from point-estimate attributions. Qualitative results on measured signals further suggest that these patterns generalise beyond the synthetic training distribution. The framework is domain-agnostic and can be applied to any BNN paired with a Lipschitz-continuous attribution operator.

Chenhao Xu, Chang-Tsun Li, Yongjian Hu et al.

Video instance segmentation, also known as multi-object tracking and segmentation, is an emerging computer vision research area introduced in 2019, aiming at detecting, segmenting, and tracking instances in videos simultaneously. By tackling the video instance segmentation tasks through effective analysis and utilization of visual information in videos, a range of computer vision-enabled applications (e.g., human action recognition, medical image processing, autonomous vehicle navigation, surveillance, etc) can be implemented. As deep-learning techniques take a dominant role in various computer vision areas, a plethora of deep-learning-based video instance segmentation schemes have been proposed. This survey offers a multifaceted view of deep-learning schemes for video instance segmentation, covering various architectural paradigms, along with comparisons of functional performance, model complexity, and computational overheads. In addition to the common architectural designs, auxiliary techniques for improving the performance of deep-learning models for video instance segmentation are compiled and discussed. Finally, we discuss a range of major challenges and directions for further investigations to help advance this promising research field.

Chenhao Xu, Chang-Tsun Li, Chee Peng Lim et al.

Due to its deficiency in prior knowledge (inductive bias), Vision Transformer (ViT) requires pre-training on large-scale datasets to perform well. Moreover, the growing layers and parameters in ViT models impede their applicability to devices with limited computing resources. To mitigate the aforementioned challenges, this paper introduces a novel horizontally scalable vision transformer (HSViT) scheme. Specifically, a novel image-level feature embedding is introduced to ViT, where the preserved inductive bias allows the model to eliminate the need for pre-training while outperforming on small datasets. Besides, a novel horizontally scalable architecture is designed, facilitating collaborative model training and inference across multiple computing devices. The experimental results depict that, without pre-training, HSViT achieves up to 10% higher top-1 accuracy than state-of-the-art schemes on small datasets, while providing existing CNN backbones up to 3.1% improvement in top-1 accuracy on ImageNet. The code is available at https://github.com/xuchenhao001/HSViT.

Zheyang Huang, Jagannath Aryal, Saeid Nahavandi et al.

Cross-view geo-localization determines the location of a query image, captured by a drone or ground-based camera, by matching it to a geo-referenced satellite image. While traditional approaches focus on image-level localization, many applications, such as search-and-rescue, infrastructure inspection, and precision delivery, demand object-level accuracy. This enables users to prompt a specific object with a single click on a drone image to retrieve precise geo-tagged information of the object. However, variations in viewpoints, timing, and imaging conditions pose significant challenges, especially when identifying visually similar objects in extensive satellite imagery. To address these challenges, we propose an Object-level Cross-view Geo-localization Network (OCGNet). It integrates user-specified click locations using Gaussian Kernel Transfer (GKT) to preserve location information throughout the network. This cue is dually embedded into the feature encoder and feature matching blocks, ensuring robust object-specific localization. Additionally, OCGNet incorporates a Location Enhancement (LE) module and a Multi-Head Cross Attention (MHCA) module to adaptively emphasize object-specific features or expand focus to relevant contextual regions when necessary. OCGNet achieves state-of-the-art performance on a public dataset, CVOGL. It also demonstrates few-shot learning capabilities, effectively generalizing from limited examples, making it suitable for diverse applications (https://github.com/ZheyangH/OCGNet).

Masoud Abdi, Ehsan Abbasnejad, Chee Peng Lim et al.

Tremendous amounts of expensive annotated data are a vital ingredient for state-of-the-art 3d hand pose estimation. Therefore, synthetic data has been popularized as annotations are automatically available. However, models trained only with synthetic samples do not generalize to real data, mainly due to the gap between the distribution of synthetic and real data. In this paper, we propose a novel method that seeks to predict the 3d position of the hand using both synthetic and partially-labeled real data. Accordingly, we form a shared latent space between three modalities: synthetic depth image, real depth image, and pose. We demonstrate that by carefully learning the shared latent space, we can find a regression model that is able to generalize to real data. As such, we show that our method produces accurate predictions in both semi-supervised and unsupervised settings. Additionally, the proposed model is capable of generating novel, meaningful, and consistent samples from all of the three domains. We evaluate our method qualitatively and quantitively on two highly competitive benchmarks (i.e., NYU and ICVL) and demonstrate its superiority over the state-of-the-art methods. The source code will be made available at https://github.com/masabdi/LSPS.

Ali Jafarizadeh, Shadi Farabi Maleki, Parnia Pouya et al.

Retinopathy of prematurity (ROP) is a severe condition affecting premature infants, leading to abnormal retinal blood vessel growth, retinal detachment, and potential blindness. While semi-automated systems have been used in the past to diagnose ROP-related plus disease by quantifying retinal vessel features, traditional machine learning (ML) models face challenges like accuracy and overfitting. Recent advancements in deep learning (DL), especially convolutional neural networks (CNNs), have significantly improved ROP detection and classification. The i-ROP deep learning (i-ROP-DL) system also shows promise in detecting plus disease, offering reliable ROP diagnosis potential. This research comprehensively examines the contemporary progress and challenges associated with using retinal imaging and artificial intelligence (AI) to detect ROP, offering valuable insights that can guide further investigation in this domain. Based on 89 original studies in this field (out of 1487 studies that were comprehensively reviewed), we concluded that traditional methods for ROP diagnosis suffer from subjectivity and manual analysis, leading to inconsistent clinical decisions. AI holds great promise for improving ROP management. This review explores AI's potential in ROP detection, classification, diagnosis, and prognosis.

Camilo Gonzalez Arango, Houshyar Asadi, Mohammad Reza Chalak Qazani et al.

Motion Cueing Algorithms (MCAs) encode the movement of simulated vehicles into movement that can be reproduced with a motion simulator to provide a realistic driving experience within the capabilities of the machine. This paper introduces a novel learning-based MCA for serial robot-based motion simulators. Building on the differentiable predictive control framework, the proposed method merges the advantages of Nonlinear Model Predictive Control (NMPC) - notably nonlinear constraint handling and accurate kinematic modeling - with the computational efficiency of machine learning. By shifting the computational burden to offline training, the new algorithm enables real-time operation at high control rates, thus overcoming the key challenge associated with NMPC-based motion cueing. The proposed MCA incorporates a nonlinear joint-space plant model and a policy network trained to mimic NMPC behavior while accounting for joint acceleration, velocity, and position limits. Simulation experiments across multiple motion cueing scenarios showed that the proposed algorithm performed on par with a state-of-the-art NMPC-based alternative in terms of motion cueing quality as quantified by the RMSE and correlation coefficient with respect to reference signals. However, the proposed algorithm was on average 400 times faster than the NMPC baseline. In addition, the algorithm successfully generalized to unseen operating conditions, including motion cueing scenarios on a different vehicle and real-time physics-based simulations.

Yinsong Chen, Samson S. Yu, Zhong Li et al.

In recent years, inconsistency in Bayesian deep learning has attracted significant attention. Tempered or generalized posterior distributions are frequently employed as direct and effective solutions. Nonetheless, the underlying mechanisms and the effectiveness of generalized posteriors remain active research topics. In this work, we interpret posterior tempering as a correction for model misspecification via adjustments to the joint probability, and as a recalibration of priors by reducing aleatoric uncertainty. We also introduce the generalized Laplace approximation, which requires only a simple modification to the Hessian calculation of the regularized loss and provides a flexible and scalable framework for high-quality posterior inference. We evaluate the proposed method on state-of-the-art neural networks and real-world datasets, demonstrating that the generalized Laplace approximation enhances predictive performance.

Salim Rezvani, Farhad Pourpanah, Chee Peng Lim et al.

This paper presents a review on methods for class-imbalanced learning with the Support Vector Machine (SVM) and its variants. We first explain the structure of SVM and its variants and discuss their inefficiency in learning with class-imbalanced data sets. We introduce a hierarchical categorization of SVM-based models with respect to class-imbalanced learning. Specifically, we categorize SVM-based models into re-sampling, algorithmic, and fusion methods, and discuss the principles of the representative models in each category. In addition, we conduct a series of empirical evaluations to compare the performances of various representative SVM-based models in each category using benchmark imbalanced data sets, ranging from low to high imbalanced ratios. Our findings reveal that while algorithmic methods are less time-consuming owing to no data pre-processing requirements, fusion methods, which combine both re-sampling and algorithmic approaches, generally perform the best, but with a higher computational load. A discussion on research gaps and future research directions is provided.

Farhad Pourpanah, Chee Peng Lim, Ali Etemad et al.

Most semi-supervised learning (SSL) models entail complex structures and iterative training processes as well as face difficulties in interpreting their predictions to users. To address these issues, this paper proposes a new interpretable SSL model using the supervised and unsupervised Adaptive Resonance Theory (ART) family of networks, which is denoted as SSL-ART. Firstly, SSL-ART adopts an unsupervised fuzzy ART network to create a number of prototype nodes using unlabeled samples. Then, it leverages a supervised fuzzy ARTMAP structure to map the established prototype nodes to the target classes using labeled samples. Specifically, a one-to-many (OtM) mapping scheme is devised to associate a prototype node with more than one class label. The main advantages of SSL-ART include the capability of: (i) performing online learning, (ii) reducing the number of redundant prototype nodes through the OtM mapping scheme and minimizing the effects of noisy samples, and (iii) providing an explanation facility for users to interpret the predicted outcomes. In addition, a weighted voting strategy is introduced to form an ensemble SSL-ART model, which is denoted as WESSL-ART. Every ensemble member, i.e., SSL-ART, assigns {\color{black}a different weight} to each class based on its performance pertaining to the corresponding class. The aim is to mitigate the effects of training data sequences on all SSL-ART members and improve the overall performance of WESSL-ART. The experimental results on eighteen benchmark data sets, three artificially generated data sets, and a real-world case study indicate the benefits of the proposed SSL-ART and WESSL-ART models for tackling pattern classification problems.

Farhad Pourpanah, Moloud Abdar, Yuxuan Luo et al.

Generalized zero-shot learning (GZSL) aims to train a model for classifying data samples under the condition that some output classes are unknown during supervised learning. To address this challenging task, GZSL leverages semantic information of the seen (source) and unseen (target) classes to bridge the gap between both seen and unseen classes. Since its introduction, many GZSL models have been formulated. In this review paper, we present a comprehensive review on GZSL. Firstly, we provide an overview of GZSL including the problems and challenges. Then, we introduce a hierarchical categorization for the GZSL methods and discuss the representative methods in each category. In addition, we discuss the available benchmark data sets and applications of GZSL, along with a discussion on the research gaps and directions for future investigations.

Farhad Pourpanah, Ran Wang, Chee Peng Lim et al.

The Artificial Fish Swarm Algorithm (AFSA) is inspired by the ecological behaviors of fish schooling in nature, viz., the preying, swarming and following behaviors. Owing to a number of salient properties, which include flexibility, fast convergence, and insensitivity to the initial parameter settings, the family of AFSA has emerged as an effective Swarm Intelligence (SI) methodology that has been widely applied to solve real-world optimization problems. Since its introduction in 2002, many improved and hybrid AFSA models have been developed to tackle continuous, binary, and combinatorial optimization problems. This paper aims to present a concise review of the continuous AFSA, encompassing the original ASFA, its improvements and hybrid models, as well as their associated applications. We focus on articles published in high-quality journals since 2013. Our review provides insights into AFSA parameters modifications, procedures and sub-functions. The main reasons for these enhancements and the comparison results with other hybrid methods are discussed. In addition, hybrid, multi-objective and dynamic AFSA models that have been proposed to solve continuous optimization problems are elucidated. We also analyse possible AFSA enhancements and highlight future research directions for advancing AFSA-based models.

Thanh Thi Nguyen, Ngoc Duy Nguyen, Peter Vamplew et al.

This paper introduces a new scalable multi-objective deep reinforcement learning (MODRL) framework based on deep Q-networks. We develop a high-performance MODRL framework that supports both single-policy and multi-policy strategies, as well as both linear and non-linear approaches to action selection. The experimental results on two benchmark problems (two-objective deep sea treasure environment and three-objective Mountain Car problem) indicate that the proposed framework is able to find the Pareto-optimal solutions effectively. The proposed framework is generic and highly modularized, which allows the integration of different deep reinforcement learning algorithms in different complex problem domains. This therefore overcomes many disadvantages involved with standard multi-objective reinforcement learning methods in the current literature. The proposed framework acts as a testbed platform that accelerates the development of MODRL for solving increasingly complicated multi-objective problems.

Thanh Nguyen, Chee Peng Lim, Ngoc Duy Nguyen et al.

Situation awareness (SA) is an important constituent in human information processing and essential in pilots' decision-making processes. Acquiring and maintaining appropriate levels of SA is critical in aviation environments as it affects all decisions and actions taking place in flights and air traffic control. This paper provides an overview of recent measurement models and approaches to establishing and enhancing SA in aviation environments. Many aspects of SA are examined including the classification of SA techniques into six categories, and different theoretical SA models from individual, to shared or team, and to distributed or system levels. Quantitative and qualitative perspectives pertaining to SA methods and issues of SA for unmanned vehicles are also addressed. Furthermore, future research directions regarding SA assessment approaches are raised to deal with shortcomings of the existing state-of-the-art methods in the literature.