Physical Sciences and Mathematics Commons^™

Rhythmedge: Enabling Contactless Heart Rate Estimation On The Edge, Zahid Hasan, Emon Dey, Sreenivasan Ramasamy Ramamurthy, Nirmalya Roy, Archan Misra

Research Collection School Of Computing and Information Systems

The primary contribution of this paper is designing and prototyping a real-time edge computing system, RhythmEdge, that is capable of detecting changes in blood volume from facial videos (Remote Photoplethysmography; rPPG), enabling cardio-vascular health assessment instantly. The benefits of RhythmEdge include non-invasive measurement of cardiovascular activity, real-time system operation, inexpensive sensing components, and computing. RhythmEdge captures a short video of the skin using a camera and extracts rPPG features to estimate the Photoplethysmography (PPG) signal using a multi-task learning framework while offloading the edge computation. In addition, we intelligently apply a transfer learning approach to the multi-task learning framework to …

Multimodal Zero-Shot Hateful Meme Detection, Jiawen Zhu, Roy Ka-Wei Lee, Wen Haw Chong

Research Collection School Of Computing and Information Systems

Facebook has recently launched the hateful meme detection challenge, which garnered much attention in academic and industry research communities. Researchers have proposed multimodal deep learning classification methods to perform hateful meme detection. While the proposed methods have yielded promising results, these classification methods are mostly supervised and heavily rely on labeled data that are not always available in the real-world setting. Therefore, this paper explores and aims to perform hateful meme detection in a zero-shot setting. Working towards this goal, we propose Target-Aware Multimodal Enhancement (TAME), which is a novel deep generative framework that can improve existing hateful meme classification …

A Simple Data Mixing Prior For Improving Self-Supervised Learning, Sucheng Ren, Huiyu Wang, Zhengqi Gao, Shengfeng He, Alan Yuille, Yuyin Zhou, Cihang Xie

Research Collection School Of Computing and Information Systems

Data mixing (e.g., Mixup, Cutmix, ResizeMix) is an essential component for advancing recognition models. In this paper, we focus on studying its effectiveness in the self-supervised setting. By noticing the mixed images that share the same source images are intrinsically related to each other, we hereby propose SDMP, short for Simple Data Mixing Prior, to capture this straightforward yet essential prior, and position such mixed images as additional positive pairs to facilitate self-supervised representation learning. Our experiments verify that the proposed SDMP enables data mixing to help a set of self-supervised learning frameworks (e.g., MoCo) achieve better accuracy and out-of-distribution …

Co-Advise: Cross Inductive Bias Distillation, Sucheng Ren, Zhengqi Gao, Tiany Hua, Zihui Xue, Yonglong Tian, Shengfeng He, Hang Zhao

Research Collection School Of Computing and Information Systems

The inductive bias of vision transformers is more relaxed that cannot work well with insufficient data. Knowledge distillation is thus introduced to assist the training of transformers. Unlike previous works, where merely heavy convolution-based teachers are provided, in this paper, we delve into the influence of models inductive biases in knowledge distillation (e.g., convolution and involution). Our key observation is that the teacher accuracy is not the dominant reason for the student accuracy, but the teacher inductive bias is more important. We demonstrate that lightweight teachers with different architectural inductive biases can be used to co-advise the student transformer with …

Do-Gan: A Double Oracle Framework For Generative Adversarial Networks, Aye Phyu Phye Aung, Xinrun Wang, Runsheng Yu, Bo An, Senthilnath Jayavelu, Xiaoli Li

Research Collection School Of Computing and Information Systems

In this paper, we propose a new approach to train Gen-erative Adversarial Networks (GANs) where we deploy a double-oracle framework using the generator and discrim-inator oracles. GAN is essentially a two-player zero-sum game between the generator and the discriminator. Training GANs is challenging as a pure Nash equilibrium may not exist and even finding the mixed Nash equilibrium is difficult as GANs have a large-scale strategy space. In DO-GAN, we extend the double oracle framework to GANs. We first generalize the players' strategies as the trained models of generator and discriminator from the best response or-acles. We then compute the …

Deep One-Class Classification Via Interpolated Gaussian Descriptor, Yuanhong Chen, Yu Tian, Guansong Pang, Gustavo Carneiro

Research Collection School Of Computing and Information Systems

One-class classification (OCC) aims to learn an effective data description to enclose all normal training samples and detect anomalies based on the deviation from the data description. Current state-of-the-art OCC models learn a compact normality description by hyper-sphere minimisation, but they often suffer from overfitting the training data, especially when the training set is small or contaminated with anomalous samples. To address this issue, we introduce the interpolated Gaussian descriptor (IGD) method, a novel OCC model that learns a one-class Gaussian anomaly classifier trained with adversarially interpolated training samples. The Gaussian anomaly classifier differentiates the training samples based on their …

A Practical Comparison Of Quantum And Classical Leaderless Consensus, Paul Robert Griffin, Dimple Mevada

Research Collection School Of Computing and Information Systems

Quantum computing is coming of age and being explored in many business areas for either solving difficult problems or improving business processes. Distributed ledger technology (DLT) is now embedded in many businesses and continues to mature. Consensus, at the heart of DLTs, has practical scaling issues and, as we move into needing bigger datasets, bigger networks and more security, the problem is ever increasing. Consensus agreement is a non-deterministic problem which is a good match to quantum computers due to the probabilistic nature of quantum phenomena. In this paper, we show that quantum nodes entangled in a variety of network …

Cross-Lingual Adaptation For Recipe Retrieval With Mixup, Bin Zhu, Chong-Wah Ngo, Jingjing Chen, Wing-Kwong Chan

Research Collection School Of Computing and Information Systems

Cross-modal recipe retrieval has attracted research attention in recent years, thanks to the availability of large-scale paired data for training. Nevertheless, obtaining adequate recipe-image pairs covering the majority of cuisines for supervised learning is difficult if not impossible. By transferring knowledge learnt from a data-rich cuisine to a data-scarce cuisine, domain adaptation sheds light on this practical problem. Nevertheless, existing works assume recipes in source and target domains are mostly originated from the same cuisine and written in the same language. This paper studies unsupervised domain adaptation for image-to-recipe retrieval, where recipes in source and target domains are in different …

Reinforcement Learning-Based Interactive Video Search, Zhixin Ma, Jiaxin Wu, Zhijian Hou, Chong-Wah Ngo

Research Collection School Of Computing and Information Systems

Despite the rapid progress in text-to-video search due to the advancement of cross-modal representation learning, the existing techniques still fall short in helping users to rapidly identify the search targets. Particularly, in the situation that a system suggests a long list of similar candidates, the user needs to painstakingly inspect every search result. The experience is frustrated with repeated watching of similar clips, and more frustratingly, the search targets may be overlooked due to mental tiredness. This paper explores reinforcement learning-based (RL) searching to relieve the user from the burden of brute force inspection. Specifically, the system maintains a graph …

Learnings From A Pilot Hybrid Question Answering System: Cqas: Case Study Based On A Singapore Government Agency's Customer Service Centre, Hui Shan Lee, Shankararaman, Venky, Eng Lieh Ouh

Research Collection School Of Computing and Information Systems

The Singapore Government first released their digital government blueprint in 2018 with the key message for all their agencies to be "digital to the core and served with heart". With this push, agencies are moving towards human-centric digital services, especially for individual citizens. During COVID-19, Singapore government agencies introduced many COVID-19 digital initiatives resulting in more incoming inquiries from citizens to respective agencies. This surge in inquiries created the challenge on the agencies' end to meet service level agreements. One widely adopted solution is the use of chatbot technology that directly interfaces with the customer. However, several organisations have faced …

Generative Flows With Invertible Attentions, Rhea Sanjay Sukthanker, Zhiwu Huang, Suryansh Kumar, Radu Timofte, Luc Van Gool

Research Collection School Of Computing and Information Systems

Flow-based generative models have shown an excellent ability to explicitly learn the probability density function of data via a sequence of invertible transformations. Yet, learning attentions in generative flows remains understudied, while it has made breakthroughs in other domains. To fill the gap, this paper introduces two types of invertible attention mechanisms, i.e., map-based and transformer-based attentions, for both unconditional and conditional generative flows. The key idea is to exploit a masked scheme of these two attentions to learn long-range data dependencies in the context of generative flows. The masked scheme allows for invertible attention modules with tractable Jacobian determinants, …

Officers: Operational Framework For Intelligent Crime-And-Emergency Response Scheduling, Jonathan David Chase, Siong Thye Goh, Tran Phong, Hoong Chuin Lau

Research Collection School Of Computing and Information Systems

In the quest to achieve better response times in dense urban environments, law enforcement agencies are seeking AI-driven planning systems to inform their patrol strategies. In this paper, we present a framework, OFFICERS, for deployment planning that learns from historical data to generate deployment schedules on a daily basis. We accurately predict incidents using ST-ResNet, a deep learning technique that captures wide-ranging spatio-temporal dependencies, and solve a large-scale optimization problem to schedule deployment, significantly improving its scalability through a simulated annealing solver. Methodologically, our approach outperforms our previous works where prediction was done using Generative Adversarial Networks, and optimization was …

A Self-Consistent Model For Brown Dwarf Populations, R. E. Ryan Jr., P. Thorman, C. Aganze, A. J. Burgasser, S. H. Cohen, N. P. Hathi, Benne Holwerda, N. Pirzkal, R. A. Windhorst

Faculty and Staff Scholarship

We present a self-consistent model of the Milky Way to reproduce the observed distributions (spectral type, absolute J-band magnitude, effective temperature) and total velocity dispersion of brown dwarfs. For our model, we adopt parametric forms for the star formation history and initial-mass function, published evolutionary models, and theoretical age–velocity relations. Using standard Markov Chain Monte Carlo methods, we derive a power-law index of the initial-mass function of α = −0.71 ± 0.11, which is an improvement over previous studies. We consider a gamma-function form for the star formation history, though we find that this complex model is only slightly …

On Three-Dimensional Rotating Viscoelastic Jets In The Giesekus Model, Daniel N. Riahi, Saulo Orizaga

School of Mathematical & Statistical Sciences Faculty Publications

We investigate three-dimensional nonlinear rotating viscoelastic curved jets in the presence of gravity force. Applying the Giesekus model for the viscoelastic stress parts of the jet flow system and using perturbation methods with a consistent scaling, a relatively simple system of equations with realistic three-dimensional centerlines is developed. We determine numerically the relevant solution quantities of the model in terms of the radius, speed, tensile force, stretching rate, strain rate and the jet centerline versus arc length and for different parameter values associated with gravity, viscosity, rotation, surface tension and viscoelasticity. Considering the jet flow system in full 3-dimensions and …

Mangrove Growth Promotion By Endophytic Actinobacteria And Seaweed Extract, Ameera Khalfan Salem Alkaabi

Theses

In this study, I aimed to determine the impact of the application of a commercial seaweed extract (SWE) bio-stimulant and endophytic actinobacterial isolates on growth performance and endogenous hormonal levels of mangroves. Therefore, I isolated endophytic plant growth promoting (PGP) actinobacteria (PGPA) from mangrove roots; and evaluated their potential as biological inoculants on mangrove seedlings under greenhouse and open-field nursery conditions. Seven salt tolerant isolates had the ability to produce different levels of in vitro plant growth regulators (PGRs) and 1-aminocyclopropane-1-carboxylic acid (ACC) deaminase (ACCD), and to solubilize phosphorus. Accordingly, only one isolate, Streptomyces tubercidicus UAE1 (St), was …

Determining Knowledge From Student Performance Prediction Using Machine Learning, Wala El Rashied Mohamed

Theses

Recent years have seen a rapid development in the field of educational data mining (EDM), enhancing the ability to trace student knowledge. Data from intelligent tutoring systems (ITS) have been analyzed and interpreted by multiple researchers seeking to measure students’ knowledge as it evolves. Human nature, as well as other factors, makes it difficult to determine whether or not students are knowledgeable. This thesis sets out to examine the level of students’ knowledge by predicting their current and future academic performance based on records of their historical interactions. By restructuring data and considering a student perspective, we can gain insight …

Weakly-Supervised Tumor Purity Prediction From Frozen H&E Stained Slides, Matthew Brendel, Vanesa Getseva, Majd Al Assaad, Michael Sigouros, Alexandros Sigaras, Troy Kane, Pegah Khosravi, Juan Miguel Mosquera, Olivier Elemento, Iman Hajirasouliha

Publications and Research

Background

Estimating tumor purity is especially important in the age of precision medicine. Purity estimates have been shown to be critical for correction of tumor sequencing results, and higher purity samples allow for more accurate interpretations from next-generation sequencing results. Molecular-based purity estimates using computational approaches require sequencing of tumors, which is both time-consuming and expensive.

Methods

Here we propose an approach, weakly-supervised purity (wsPurity), which can accurately quantify tumor purity within a digitally captured hematoxylin and eosin (H&E) stained histological slide, using several types of cancer from The Cancer Genome Atlas (TCGA) as a proof-of-concept.

Findings

Our model predicts …

Emissions Baseline Report For The Agriculture Sector In Western Australia, Mandy Curnow, Christophe D'Abbadie, Robert French, Jackie Bucat, Dario Stefanelli, Neil Lantzke, Valeria Almeida Lima, Jess Austen, Lee Chester, Kate Pritchett, Millie Brady, Dylan Bennett

Climate Science Published Reports

The Intergovernmental Panel on Climate Change Sixth Assessment Report found that climate impacts are appearing earlier and are more severe than expected. The world faces multiple unavoidable climate hazards over the next two decades with global warming of 1.5ºC and accelerated action is required to avoid mortality and loss of biodiversity and infrastructure.

Achieving Climate Resilience requires both climate adaptation and mitigation of Greenhouse Gases which are accelerating climate change. Western Australia is committed to building resilience of the agriculture industry and as part of that is focussed on reducing greenhouse gas (GHG) emissions at both an industry and government …

First Joint Observation By The Underground Gravitational-Wave Detector Kagra With Geo 600, R. Abbott, H. Abe, Fausto Acernese, K. Ackley, N. Adhikari, Michael G. Benjamin, Teviet Creighton, Mario C. Diaz, Francisco Llamas, Soma Mukherjee, Gaukhar Nurbek, Volker Quetschke, Wenhui Wang

Physics & Astronomy Faculty Publications

We report the results of the first joint observation of the KAGRA detector with GEO 600. KAGRA is a cryogenic and underground gravitational-wave detector consisting of a laser interferometer with 3 km arms, located in Kamioka, Gifu, Japan. GEO 600 is a British–German laser interferometer with 600 m arms, located near Hannover, Germany. GEO 600 and KAGRA performed a joint observing run from April 7 to 20, 2020. We present the results of the joint analysis of the GEO–KAGRA data for transient gravitational-wave signals, including the coalescence of neutron-star binaries and generic unmodeled transients. We also perform dedicated searches for …

Suppressing Disorder-Order Phase Transition Of Gd2zr2o7 Pyrochlore By Dy3+ Doping And Their Impact On Luminescence, Santosh K. Gupta, S. Nigam, Jose P. Zuniga, Yuanbing Mao

School of Integrative Biological & Chemical Sciences (Formerly Dept. of Chemistry)

The functionality of the Gd2Zr2O7 matrix is dictated by ordering at the cationic lattice (pyrochlore or defective fluorite) and associated local manifestations. In this work, we have synthesized nano and bulk undoped Gd2Zr2O7 (GZO) and Gd2Zr2O7:Dy3+ (GZOD) by molten salt synthesis adopting a crystal engineering approach and investigated their impact on photoluminescence properties. DFT calculations along with the higher intensity of yellow emission in the bulk GZOD confirm that the presence of Dy3+ is energetically favorable in defect fluorite structure compared to ordered pyrochlore counterpart. Increase/adjustments in the Zr–O bond distances in Dy3+ surroundings facilitate the release of local strain/stress …

A Unified View Of A Human Digital Twin, Michael Miller, Emily Spatz

Faculty Publications

The term human digital twin has recently been applied in many domains, including medical and manufacturing. This term extends the digital twin concept, which has been illustrated to provide enhanced system performance as it combines system models and analyses with real-time measurements for an individual system to improve system maintenance. Human digital twins have the potential to change the practice of human system integration as these systems employ real-time sensing and feedback to tightly couple measurements of human performance, behavior, and environmental influences throughout a product’s life cycle to human models to improve system design and performance. However, as this …

Runtime Energy Savings Based On Machine Learning Models For Multicore Applications, Vaibhav Sundriyal, Masha Sosonkina

Electrical & Computer Engineering Faculty Publications

To improve the power consumption of parallel applications at the runtime, modern processors provide frequency scaling and power limiting capabilities. In this work, a runtime strategy is proposed to maximize energy savings under a given performance degradation. Machine learning techniques were utilized to develop performance models which would provide accurate performance prediction with change in operating core-uncore frequency. Experiments, performed on a node (28 cores) of a modern computing platform showed significant energy savings of as much as 26% with performance degradation of as low as 5% under the proposed strategy compared with the execution in the unlimited power case.

Coded Distributed Function Computation, Pedro J. Soto

Dissertations, Theses, and Capstone Projects

A ubiquitous problem in computer science research is the optimization of computation on large data sets. Such computations are usually too large to be performed on one machine and therefore the task needs to be distributed amongst a network of machines. However, a common problem within distributed computing is the mitigation of delays caused by faulty machines. This can be performed by the use of coding theory to optimize the amount of redundancy needed to handle such faults. This problem differs from classical coding theory since it is concerned with the dynamic coded computation on data rather than just statically …

Why, New York City? Gauging The Quality Of Life Through The Thoughts Of Tweeters, Sheryl Williams

Dissertations, Theses, and Capstone Projects

As a resource for social data, Twitter’s platform has been used to measure the quality of life through sentiment analysis. This capstone project explores another methodological technique—querying Twitter data around specific keyword terms to determine dominant topics, word patterns, and sentiment leanings in a geographical area. Focusing on New York City and Los Angeles for comparative analysis, the keyword term “why” will be used to build a Python analysis around topic modeling and sentiment analysis. Using this approach, the analysis reveals social and cultural differences, the overall sentiment of tweets, and subjects of interest to tweeters.

GitHub Repository for all …

Collective Behavior Of Dissipatively-Coupled Photonic Oscillator Networks, Jiajie Ding

Dissertations, Theses, and Capstone Projects

This thesis discusses the collective behavior of networks formed by photonic oscillators. The key motivations for doing this research are the interest in developing high power laser arrays on photonic chips and performing optical computing.

A Comparison Of Machine Learning Techniques For Validating Students’ Proficiency In Mathematics, Alexander Avdeev

Dissertations, Theses, and Capstone Projects

A principal goal of this project was to compare several machine learning (ML) algorithms to explore and validate math proficiency classifications based on standardized test scores. The data used in these analyses came from the 6th-grade students’ mathematics assessment records of the New York State Education Department’s Testing Program (NYSTP). Our approach was to test a number of competing machine learning (ML) algorithms for classifying students’ as proficient based on their test scores and other demographic information. Our samples were drawn from the 2016 test-taking cohort of 6th-grade students (N=156,800). Five classifiers including multinominal logistic regression (MLR), XGBoost, Tree-As, Lagrangian …

Symmetry-Inspired Analysis Of Biological Networks, Ian Leifer

Dissertations, Theses, and Capstone Projects

The description of a complex system like gene regulation of a cell or a brain of an animal in terms of the dynamics of each individual element is an insurmountable task due to the complexity of interactions and the scores of associated parameters. Recent decades brought about the description of these systems that employs network models. In such models the entire system is represented by a graph encapsulating a set of independently functioning objects and their interactions. This creates a level of abstraction that makes the analysis of such large scale system possible. Common practice is to draw conclusions about …

Heating Fire Incidents In New York City, Merissa K. Lissade

Dissertations, Theses, and Capstone Projects

If you have ever had the Citizen app downloaded on your smart phone, then you know how many alerts you receive in a day living in New York City (NYC). Citizen is a mobile app that sends real-time safety alerts based on the location of its user. In my experience having the app, I have seen many notifications of fires caused by heaters during the winter. On the morning of January 9^th, 2022, I received a notification of an accidental blaze that took the lives of 17 people from the choking smoke of a 19-story residential building in …

Biomimetic And Medical Applications Of Hollow Nanoscale Structures, Justin Fang

Dissertations, Theses, and Capstone Projects

Materials whose structure incorporates nanoscale void spaces have multiple possible uses, whether in a bulk form or as individual particles, due to the combination of high surface area ratios and nanoscale material properties. This thesis will explore a few of these possibilities, concentrating on potential biomimetic and biomedical applications, for two materials: metal- organic frameworks and superparamagnetic iron oxide nanocages.

Metal-organic frameworks consist of metal ions such as Cu2+ which have highly porous lattice structures allowing them to absorb and release guest molecules such as peptides like diphenylalanine; this stored chemical energy can be turned into kinetic energy and used …

The Zariski-Riemann Space As A Universal Model For The Birational Geometry Of A Function Field, Giovan Battista Pignatti Morano Di Custoza

Dissertations, Theses, and Capstone Projects

Given a function field $K$ over an algebraically closed field $k$, we propose to use the Zariski-Riemann space $\ZR (K/k)$ of valuation rings as a universal model that governs the birational geometry of the field extension $K/k$. More specifically, we find an exact correspondence between ad-hoc collections of open subsets of $\ZR (K/k)$ ordered by quasi-refinements and the category of normal models of $K/k$ with morphisms the birational maps. We then introduce suitable Grothendieck topologies and we develop a sheaf theory on $\ZR (K/k)$ which induces, locally at once, the sheaf theory of each normal model. Conversely, given a sheaf …