Computer Sciences Commons^™

Research@Smu: Sustainable Living, Singapore Management University

Research Collection Office of Research

Sustainable Living is one of the three key priorities of the SMU 2025 Strategy, and the University is committed to develop it into an area of cross-disciplinary strength. The articles in this booklet highlight impactful sustainability research accomplishments at SMU, which spans five broad pillars: Sustainable Business Operations; Sustainable Finance and Impact Assessment; Sustainable Ageing and Wellness; Sustainable Urban Infrastructure; and Sustainable Agro-business and Food Consumption.

Contents:

Sustainable Business Operations

• Managing the Load on Loading Bays
• Going the Last-mile
• Feeding a Growing World
• Pooling the Benefits of Sharing a Ride

Sustainable Finance and Impact Assessment

• When Going Green Becomes a …

A Path Planning Framework For Multi-Agent Robotic Systems Based On Multivariate Skew-Normal Distributions, Peter Estephan

Theses, Dissertations and Capstones

This thesis presents a path planning framework for a very-large-scale robotic (VLSR) system in an known obstacle environment, where the time-varying distributions of agents are applied to represent the multi-agent robotic system (MARS). A novel family of the multivariate skew-normal (MVSN) distributions is proposed based on the Bernoulli random field (BRF) referred to as the Bernoulli-random-field based skew-normal (BRF-SN) distribution. The proposed distributions are applied to model the agents’ distributions in an obstacle-deployed environment, where the obstacle effect is represented by a skew function and separated from the no-obstacle agents’ distributions. First, the obstacle layout is represented by a Hilbert …

Encryption And Compression Classification Of Internet Of Things Traffic, Mariam Najdat M Saleh

Browse all Theses and Dissertations

The Internet of Things (IoT) is used in many fields that generate sensitive data, such as healthcare and surveillance. Increased reliance on IoT raised serious information security concerns. This dissertation presents three systems for analyzing and classifying IoT traffic using Deep Learning (DL) models, and a large dataset is built for systems training and evaluation. The first system studies the effect of combining raw data and engineered features to optimize the classification of encrypted and compressed IoT traffic using Engineered Features Classification (EFC), Raw Data Classification (RDC), and combined Raw Data and Engineered Features Classification (RDEFC) approaches. Our results demonstrate …

Efficient Cloud-Based Ml-Approach For Safe Smart Cities, Niveshitha Niveshitha

Browse all Theses and Dissertations

Smart cities have emerged to tackle many critical problems that can thwart the overwhelming urbanization process, such as traffic jams, environmental pollution, expensive health care, and increasing energy demand. This Master thesis proposes efficient and high-quality cloud-based machine-learning solutions for efficient and sustainable smart cities environment. Different supervised machine-learning models for air quality predication (AQP) in efficient and sustainable smart cities environment is developed. For that, ML-based techniques are implemented using cloud-based solutions. For example, regression and classification methods are implemented using distributed cloud computing to forecast air execution time and accuracy of the implemented ML solution. These models are …

Contributors To Pathologic Depolarization In Myotonia Congenita, Jessica Hope Myers

Browse all Theses and Dissertations

Myotonia congenita is an inherited skeletal muscle disorder caused by loss-of-function mutation in the CLCN1 gene. This gene encodes the ClC-1 chloride channel, which is almost exclusively expressed in skeletal muscle where it acts to stabilize the resting membrane potential. Loss of this chloride channel leads to skeletal muscle hyperexcitability, resulting in involuntary muscle action potentials (myotonic discharges) seen clinically as muscle stiffness (myotonia). Stiffness affects the limb and facial muscles, though specific muscle involvement can vary between patients. Interestingly, respiratory distress is not part of this disease despite muscles of respiration such as the diaphragm muscle also carrying this …

Digital Twin For Railway: A Comprehensive Survey, Sara Ghaboura, Rahatara Ferdousi, Fedwa Laamarti, Chunsheng Yang, Abdulmotaleb El Saddik

Computer Vision Faculty Publications

Digital transformation has been prioritized in the railway industry to bring automation to railway operations. Digital Twin (DT) technology has recently gained attention in the railway industry to fulfill this goal. Contemporary researchers argue that DT can be advantageous in Railway manufacturing logistics to planning and scheduling. Although underlying technologies of DT, e.g., modelling, computer vision, and the Internet of Things, have been studied for various railway industry applications, the DT has been least explored in the context of railways. Thus, in this paper, we aim to understand the state-of-the-art of DT for railway (DTR), for advanced railway systems. Besides, …

Maple: Multi-Modal Prompt Learning, Muhammad Uzair Khattak, Hanoona Rasheed, Muhammad Maaz, Salman Khan, Fahad Shahbaz Khan

Computer Vision Faculty Publications

Pre-trained vision-language (V-L) models such as CLIP have shown excellent generalization ability to downstream tasks. However, they are sensitive to the choice of input text prompts and require careful selection of prompt templates to perform well. Inspired by the Natural Language Processing (NLP) literature, recent CLIP adaptation approaches learn prompts as the textual inputs to fine-tune CLIP for downstream tasks. We note that using prompting to adapt representations in a single branch of CLIP (language or vision) is sub-optimal since it does not allow the flexibility to dynamically adjust both representation spaces on a downstream task. In this work, we …

Retrospective On The Sensorium 2022 Competition, Konstantin F. Willeke, Paul G. Fahey, Mohammad Bashiri, Laura Hansel, Christoph Blessing, Konstantin Klemens Lurz, Max F. Burg, Santiago A. Cadena, Zhiwei Ding, Kayla Ponder, Taliah Muhammad, Saumil S. Patel, Kaiwen Deng, Yuanfang Guan, Yiqin Zhu, Kaiwen Xiao, Xiao Han, Simone Azeglio, Ulisse Ferrari, Peter Neri, Olivier Marre

Machine Learning Faculty Publications

The neural underpinning of the biological visual system is challenging to study experimentally, in particular as neuronal activity becomes increasingly nonlinear with respect to visual input. Artificial neural networks (ANNs) can serve a variety of goals for improving our understanding of this complex system, not only serving as predictive digital twins of sensory cortex for novel hypothesis generation in silico, but also incorporating bio-inspired architectural motifs to progressively bridge the gap between biological and machine vision. The mouse has recently emerged as a popular model system to study visual information processing, but no standardized large-scale benchmark to identify state-of-the-art models …

Analysis Of Ab Initio Protein Structure Prediction Methods, Maytha Alshammari, Jing He

College of Sciences Posters

Protein structure prediction produces atomic models of three-dimensional structure of a protein from its amino acid sequence. Understanding the function mechanism of proteins requires knowledge of three-dimensional structures. When developing new enzymes and drugs, it's essential to understand the structure of the target protein. In this study, we analyze models predicted using two ab initio protein structure prediction methods, trRosetta and Quark. A set of thirty protein chains was used to evaluate the effectiveness of the methods. The thirty chains were collected from Protein Data Bank (June – November, 2020). The length and the relative position of the predicted secondary …

Ml-Based Surrogates And Emulators, Tareq Alghamdi, Yaohang Li, Nobuo Sato

College of Sciences Posters

No abstract provided.

A Novel Parking Management In Smart City Vehicular Datacenters, Syed Rizvi, Susan Zehra, Steven Olariu

College of Sciences Posters

Researchers have shown that most vehicles spend the majority of their time parked in parking garages, lots, or driveways. During this time, their computing resources are unused and untapped. This has led to substantial interest in Vehicular Cloud, an area of research in which each vehicle acts as a computation node. The main difference between traditional cloud computing and vehicular cloud computing is the availability of nodes. In traditional clouds, nodes are available 24/7, while in vehicular clouds, nodes (vehicles) are only available while parked in parking lots. This creates a dynamic environment as vehicles enter and exit parking garages …

A Computational Analysis Of Hybrid Genome Assembly Strategies, Joseph Walewski

Computer Science Honors Papers

The central dogma of molecular biology states that DNA is transcribed to RNA and then translated into proteins. Since DNA is the starting material for many of biology’s macromolecules, it has been referred to as “nature’s instruction book.” The sum of all DNA in a cell is referred to as the genome, and genome sequencing is how we interpret the DNA.

Due to limitations on currently available technology, it is not possible to retrieve the entire genome in one contiguous set of data. Therefore, genome sequencing is a computer science problem as sequencing “reads” must be stitched together to obtain …

Using Open-Source To Enhance Teaching And Scholarship, Steven Clontz, Michael Black, Ricky Green, Carlos Montalvo, Rebecca Macdonald, Sean Stalley

CoTL 2023 Panel

The adoption of open-source resources (software, hardware, educational content, and more) that are freely licensed for use, sharing, repurposing, and remixing has grown dramatically in recent years, within both academia and industry. This panel features several faculty and staff who will share and discuss their experiences using open-source solutions to enhance teaching and scholarship (both SoTL and discipline research) at their institutions.

Differentially Private Stochastic Convex Optimization In (Non)-Euclidean Space Revisited, Jinyan Su, Changhong Zhao, Di Wang

Machine Learning Faculty Publications

In this paper, we revisit the problem of Differentially Private Stochastic Convex Optimization (DP-SCO) in Euclidean and general `dp spaces. Specifically, we focus on three settings that are still far from well understood: (1) DP-SCO over a constrained and bounded (convex) set in Euclidean space; (2) unconstrained DP-SCO in `dp space; (3) DP-SCO with heavy-tailed data over a constrained and bounded set in `dp space. For problem (1), for both convex and strongly convex loss functions, we propose methods whose outputs could achieve (expected) excess population risks that are only dependent on the Gaussian width of the constraint set, rather …

Scalable Quantum Edge Detection Method For D-Nisq Imaging Simulations: Use Cases From Nuclear Physics And Medical Image Computing, Emmanuel Billias, Nikos Chrisochoides

The Graduate School Posters

Edge Detection is one of the computationally intensive modules in image analysis. It is used to find important landmarks by identifying a significant change (or “edge”) between pixels and voxels. We present a hybrid Quantum Edge Detection method by improving three aspects of an existing widely referenced implementation, which for our use cases generates incomprehensible results for the type and size of images we are required to process. Our contributions are in the pre- and post-processing (i.e., classical phase) and a quantum edge detection circuit: (1) we use space- filling curves to eliminate image artifacts introduced by the image decomposition, …

Exploring Human Perception While Reading Fake And Real News Articles, Yasasi Abeysinghe, Gavindya Jayawardana, Autumn Woodson, Efe Bozkir, Enkelejda Kasneci, Andrew Duchowski, Sampath Jayarathna

College of Sciences Posters

With the increased spread of misinformation on online platforms and the popularity of AI-generated text, there is a critical need to detect human perception regarding the truthfulness of news. Users’ believability in a news item influences the reading and sharing of that news. Hence, in order to reduce the spread of fake news online, it is important to understand how users' engagement with fake and real news and users' perceived believability impact their behavioral and physiological factors. In this work, we study human eye movements based on the truthfulness of news and their perceived believability. Using the publicly available FakeNewsPerception …

Metaenhance: Metadata Quality Improvement For Electronic Theses And Dissertations, Muntabir H. Choudhury, Lamia Salsabil, Himarsha R. Jayanetti, Jian Wu

College of Sciences Posters

Metadata quality is crucial for digital objects to be discovered through digital library interfaces. Although DL systems have adopted Dublin Core to standardize metadata formats (e.g., ETD-MS v1.1¹), the metadata of digital objects may contain incomplete, inconsistent, and incorrect values [1]. Most existing frameworks to improve metadata quality rely on crowdsourced correction approaches, e.g., [2]. Such methods are usually slow and biased toward documents that are more discoverable by users. Artificial intelligence (AI) based methods can be adopted to overcome this limit by automatically detecting, correcting, and canonicalizing the metadata, featuring quick and unbiased responses to document metadata. …

X-Disetrac: Distributed Eye-Tracking With Extended Realities, Bhanuka Mahanama, Sampath Jayarathna

College of Sciences Posters

Humans use heterogeneous collaboration mediums such as in-person, online, and extended realities for day-to-day activities. Identifying patterns in viewpoints and pupillary responses (a.k.a eye-tracking data) provide informative cues on individual and collective behavior during collaborative tasks. Despite the increasing ubiquity of these different mediums, the aggregation and analysis of eye-tracking data in heterogeneous collaborative environments remain unexplored. Our study proposes X-DisETrac: Extended Distributed Eye Tracking, a versatile framework for eye tracking in heterogeneous environments. Our approach tackles the complexity by establishing a platform-agnostic communication protocol encompassing three data streams to simplify data aggregation and …

Nudyclr: Nuclear Dynamic Co-Learned Representations, Víctor Samuel Pérez-Díaz

2023 REYES Proceedings

NuCLR (Nuclear Co-Learned Representations) is a cutting-edge multi-task deep learning framework designed to predict essential nuclear observables, including binding energies, decay energies, and nuclear charge radii. As part of the REYES Mentorship Program, we investigated the application of dynamic loss weighting to further refine NuCLR’s predictive performance. Our findings indicate that while weighting strategies can enhance accuracy in specific tasks, such as binding energy prediction, they may underperform in others. Equal Weighting (EW), the original method employed by NuCLR, demonstrated consistent performance across multiple tasks, affirming its robustness. This report succinctly presents the developments and results of the mentorship program …

Programmable Software-Defined Testbed For Visible Light Uav Networks: Architecture Design And Implementation, Yue Zhang, Nan Cen

Computer Science Faculty Research & Creative Works

As of Today, There Has Been Increasing Research on Designing Optimization Algorithms and Intelligent Network Control Methods for Visible Light Unmanned Aerial Vehicles (UAV) Networks to Provide Pervasive and Broadband Connections. for Those Theoretical Analysis based Algorithms, there is an Urgent Need to Have a Visible Light UAV Network Platform that Can Help Evaluate the Proposed Algorithms in Real-World Scenarios. However, to the Best of Our Knowledge, there is Currently No Dedicated High Data Rate and Flexible Visible Light UAV Networking Prototype. to Bridge This Gap, in This Paper, We First Design a Novel Programmable Software-Defined Architecture for Visible Light …

Using Geographic Location-Based Public Health Features In Survival Analysis, Navid Seidi, Ardhendu S. Tripathy, Sajal K. Das

Computer Science Faculty Research & Creative Works

Time elapsed till an event of interest is often modeled using the survival analysis methodology, which estimates a survival score based on the input features. There is a resurgence of interest in developing more accurate prediction models for time-to-event prediction in personalized healthcare using modern tools such as neural networks. Higher quality features and more frequent observations improve the predictions for a patient, however, the impact of including a patient's geographic location-Based public health statistics on individual predictions has not been studied. This paper proposes a complementary improvement to survival analysis models by incorporating public health statistics in the input …

Towards A Domain-Agnostic Knowledge Graph-As-A-Service Infrastructure For Active Cyber Defense With Intelligent Agents, Prasad Calyam, Mayank Kejriwal, Praveen Rao, Jianlin Cheng, Weichao Wang, Linquan Bai, V. Sriram Siddhardh Nadendla, Sanjay Kumar Madria, Sajal K. Das, Rohit Chadha, Khaza Anuarul Hoque, Kannappan Palaniappan, Kiran Neupane, Roshan Lal Neupane, Sankeerth Gandhari, Mukesh Singhal, Lotfi Othmane, Meng Yu

Computer Science Faculty Research & Creative Works

Active cyber defense mechanisms are necessary to perform automated, and even autonomous operations using intelligent agents that defend against modern/sophisticated AI-inspired cyber threats (e.g., ransomware, cryptojacking, deep-fakes). These intelligent agents need to rely on deep learning using mature knowledge and should have the ability to apply this knowledge in a situational and timely manner for a given AI-inspired cyber threat. in this paper, we describe a 'domain-Agnostic knowledge graph-As-A-service' infrastructure that can support the ability to create/store domain-specific knowledge graphs for intelligent agent Apps to deploy active cyber defense solutions defending real-world applications impacted by AI-inspired cyber threats. Specifically, we …

Optimizing Stochastic Task Migration In Vehicular Edge Computing, Ankur Nahar, Debasis Das, Sajal K. Das

Computer Science Faculty Research & Creative Works

The performance of vehicular edge computing (VEC) depends on the effective optimization of task offloading. However, uneven distribution of vehicular traffic, rapidly changing network conditions, and stochastic nature of vehicular networks motivate us to innovate approaches to efficient resource management while maintaining system's stability. to address these challenges, we propose a novel queue length-Based stochastic task migration strategy that leverages model predictive control (MPC) and Lyapunov optimization techniques. Our approach employs the queue length at the edge node as the criterion for offloading decisions. the MPC controller dynamically allocates the processing power and bandwidth resources to vehicles based on their …

Identifying Gendered Language, Shweta Soundararajan, Sarah Jane Delany

Academic Posters Collection

Gendered language refers to the use of words that indicate the gender of an individual. It can be explicit, where the gender is directly implied by the specific words used (e.g., mother, she, man), or it can be implicit, where societal roles and behaviors convey a person's gender. For example, expectations that women display communal traits (e.g., affectionate, caring, gentle) and men display agentic traits (e.g., assertive, competitive, decisive). The presence of gendered language in natural language processing (NLP) systems can reinforce gender stereotypes and bias. Our work introduces an approach to creating gendered language datasets using ChatGPT. These datasets …

Improve Engagement With Full Labs And Motivated Students: Interactive Labs Via Low Stakes Assessment, Susan Mckeever, Patricia O'Byrne, Amanda O'Farrell

Academic Posters Collection

Poor engagement and attendance is an endemic problem at third level, particularly post covid. Our approach shows how the use of regular in-lab assessment and challenges can dramatically increase student participation and learning. Using three case studies, we demonstrate how we have successfully used this low-stakes assessment approach to improve student outcomes, across a range of modules.

Detecting Patches On Road Pavement Images Acquired With 3d Laser Sensors Using Object Detection And Deep Learning, Ibrahim Hassan Syed, Dympna O'Sullivan, Susan Mckeever, David Power, Ray Mcgowan, Kieran Feighan

Academic Posters Collection

Regular pavement inspections are key to good road maintenance and road defect corrections. Advanced pavement inspection systems such as LCMS (Laser Crack Measurement System) can automatically detect the presence of different defects using 3D lasers. However, such systems still require manual involvement to complete the detection of pavement defects. This work proposes an automatic patch detection system using an object detection technique. Results show that the object detection model can successfully detect patches inside LCMS images and suggest that the proposed approach could be integrated into the existing pavement inspection systems.

Towards Accommodating Gerunds Within The Sign Language Lexicon, Zaid Mohammed, Irene Murtagh

Academic Posters Collection

This work is part of ongoing research work that focuses on the linguistic analysis and computational description of five different Sign Languages (SLs) namely Irish Sign Language (ISL), Flemish Sign Language (VGT), Dutch Sign Language (NGT), Spanish Sign Language (LSE), and British Sign Language (BSL) as part of the SignON project. This work will be leveraged to inform the development of SL lexicon entries for a Sign Language Machine Translation (SLMT) system. In particular, this research focuses on ISL. We investigate the existence of constructions similar to or equivalent in functionality to gerunds in spoken language, in particular, English. The …

The Proof Is In The Pudding – Using Perceived Stress To Measure Short-Term Impact In Initiatives To Enhance Gender Balance In Computing Education, Alina Berry, Sarah Jane Delany

Academic Posters Collection

The problem of gender imbalance in computing higher education has forced academics and professionals to implement a wide range of initiatives. Many initiatives use recruitment or retention numbers as their most obvious evidence of impact. This type of evidence of impact is, however, more resource heavy to obtain, as well as often requires a longitudinal approach. There are many shorter term initiatives that use other ways to measure their success.

First, this poster presents with a review of existing evaluation measures in interventions to recruit and retain women in computing education across the board. Three main groups of evaluation come …

Artificial Emotional Intelligence In Socially Assistive Robots, Hojjat Abdollahi

Electronic Theses and Dissertations

Artificial Emotional Intelligence (AEI) bridges the gap between humans and machines by demonstrating empathy and affection towards each other. This is achieved by evaluating the emotional state of human users, adapting the machine’s behavior to them, and hence giving an appropriate response to those emotions. AEI is part of a larger field of studies called Affective Computing. Affective computing is the integration of artificial intelligence, psychology, robotics, biometrics, and many more fields of study. The main component in AEI and affective computing is emotion, and how we can utilize emotion to create a more natural and productive relationship between humans …

Lightesd: Fully-Automated And Lightweight Anomaly Detection Framework For Edge Computing, Ronit Das, Tie (Tony) T. Luo

Computer Science Faculty Research & Creative Works

Anomaly Detection is Widely Used in a Broad Range of Domains from Cybersecurity to Manufacturing, Finance, and So On. Deep Learning based Anomaly Detection Has Recently Drawn Much Attention Because of its Superior Capability of Recognizing Complex Data Patterns and Identifying Outliers Accurately. However, Deep Learning Models Are Typically Iteratively Optimized in a Central Server with Input Data Gathered from Edge Devices, and Such Data Transfer between Edge Devices and the Central Server Impose Substantial overhead on the Network and Incur Additional Latency and Energy Consumption. to overcome This Problem, We Propose a Fully Automated, Lightweight, Statistical Learning based Anomaly …