Eliathamby Ambikairajah's research while affiliated with UNSW Sydney and other places

Publications (377)

Article
Full-text available
People perceive emotions via multiple cues, predominantly speech and visual cues, and a number of emotion recognition systems utilize both audio and visual cues. Moreover, the perception of static aspects of emotion (speaker's arousal level is high/low) and the dynamic aspects of emotion (speaker is becoming more aroused) might be perceived via dif...
Preprint
Full-text available
Biologically inspired auditory models play an important role in developing effective audio representations that can be tightly integrated into speech and audio processing systems. Current computational models of the cochlea are typically expressed in terms of systems of differential equations and do not directly lend themselves for use in computati...
Preprint
Full-text available
There is growing interest in affective computing for the representation and prediction of emotions along ordinal scales. However, the term ordinal emotion label has been used to refer to both absolute notions such as low or high arousal, as well as relation notions such as arousal is higher at one instance compared to another. In this paper, we int...
Article
The cochlea is a remarkable spectrum analyser with desirable properties including sharp frequency tuning and level-dependent compression and the potential advantages of incorporating these characteristics in a speech processing front-end are investigated. This paper develops a framework for an active transmission line cochlear model employing adapt...
Article
In this article, we describe and discuss the design-based approach for signal processing education at the undergraduate level at the University of New South Wales (UNSW) Sydney. The electrical engineering (EE) undergraduate curriculum at UNSW Sydney includes three dedicated signal processing courses as well as a design course that involves a major...
Article
Full-text available
During the last decade, Speech Emotion Recognition (SER) has emerged as an integral component within Human-computer Interaction (HCI) and other high-end speech processing systems. Generally, an SER system targets the speaker’s existence of varied emotions by extracting and classifying the prominent features from a preprocessed speech signal. Howeve...
Chapter
Full-text available
Affective computing is a developing interdisciplinary examination field uniting specialists and experts from different fields, from artificial intelligence, natural language processing to intellectual and sociologies. The thought behind affective computing is to give computers the aptitude of insight that will, in general, comprehend human feelings...
Chapter
Full-text available
The research community’s ever-increasing interest in studying human-computer interactions (HCI), systems deducing, and identifying a speech signal’s emotional aspects has emerged as a hot research topic. Speech Emotion Recognition (SER) has brought the development of automated and intelligent analysis of human utterances to reality. Typically, an S...
Chapter
Full-text available
Nowadays, there is a growing interest in colorizing many grayscales or black and white images dating back to before the colored camera for historical and aesthetic reasons. Image and video colorization can be applied to historical images, natural images, astronomical photography. This paper proposes a fully automated image colorization using a deep...
Chapter
Full-text available
Speech is the most significant communication mode among human beings and a potential method for human-computer interaction (HCI). Being unparallel in complexity, the perception of human speech is very hard. The most crucial characteristic of speech is gender, and for the classification of gender often pitch is utilized. However, it is not a reliabl...
Chapter
In the age of digital revolution, technology plays a vital role in transforming education practices. Online courses are gaining a lot of attention and demand in recent times as universities try to manage resources and cope with increasing number of student enrolments. In this study, an existing electrical safety course comprising of various activit...
Article
Like many psychological scales, depression scales are ordinal in nature. Depression prediction from behavioural signals has so far been posed either as classification or regression problems. However, these naive approaches have fundamental issues because they are not focused on ranking, unlike ordinal regression, which is the most appropriate appro...
Article
Full-text available
Inhabited isolated areas require independent power systems called isolated microgrids. Conventional protection schemes are not suitable for isolated microgrids due to their dependence on significant fault current. This article introduces an adaptive differential feature-based protection scheme for fault detection and faulty phase identification for...
Article
Full-text available
Most research on replay detection has focused on developing a stand-alone countermeasure that runs independently of a speaker verification system by training a single spoofed model and a single genuine model for all speakers. In this paper, we explore the potential benefits of adapting the back-end of a spoofing detection system towards the claimed...
Conference Paper
Full-text available
Highly fluctuating renewable sources create enormous stability issues in microgrids. Small signal and transient stability are the major classifications in microgrid stability. For small signal studies, eigenvalue analysis based on state space modeling is the most common method. Detailed dynamic modeling of the system is required for transient analy...
Conference Paper
Full-text available
Microgrids can connect and disconnect from the grid to enable them to operate in both grid-connected and islanded modes. Stability issues are often a concern in islanded microgrids due to a decrease in system inertia. In addition, the quick response of power electronic inverters in the system creates a high rate of change in frequency, voltage and...
Conference Paper
Full-text available
Insignificant fault current in inverter-based islanded AC microgrids makes fault detection challenging. This paper introduces a voltage signal-based fault assessment method for islanded AC microgrids. It considers a new set of features such as instantaneous jumps in amplitude, phase angle and frequency of voltage signal for fault detection and clas...
Conference Paper
Full-text available
A low-complexity codec and a hardware architecture are proposed for achieving real-time compression of four-dimensional (4-D) light field (LF) signals captured from camera/lenslet arrays. The proposed system employs the 4-D extension of the two-dimensional (2-D) 8×8 approximate discrete cosine transform (ADCT) that has recently appeared in the lite...
Article
Full-text available
The auditory front-end is an integral part of a spiking neural network (SNN) when performing auditory cognitive tasks. It encodes the temporal dynamic stimulus, such as speech and audio, into an efficient, effective and reconstructable spike pattern to facilitate the subsequent processing. However, most of the auditory front-ends in current studies...
Article
First-year undergraduates may be particularly prone to experiencing difficulties with facilitating feelings of relatedness, due to the recent shift in educational environments (i.e. from high school to university), which may be unfamiliar. Therefore, the current study aimed to determine whether the implementation of a single pedagogical strategy, c...
Conference Paper
Full-text available
Almost all current speech based emotion prediction systems employ front-ends that approximately represent the distribution of frame based features over a suitable window, typically via a set of statistical functionals or the use of Bag-of-Audio-Words (BoAW) features. These front-ends are designed either by manual selection of appropriate statistica...
Preprint
Auditory front-end is an integral part of a spiking neural network (SNN) when performing auditory cognitive tasks. It encodes the temporal dynamic stimulus, such as speech and audio, into an efficient, effective and reconstructable spike pattern to facilitate the subsequent processing. However, most of the auditory front-ends in current studies hav...
Article
Low-complexity signal processing algorithms and reconfigurable digital hardware architectures are proposed for multi-depth filtering and occlusion suppression in 4-D light fields (LFs). The proposed multi-depth-pass (MDP) and multi-depth-reject (MDR) 4-D filters allow simultaneous enhancement of planar objects at multiple desired depths and attenua...
Article
This paper presents a novel method of estimating systolic blood pressure (SBP) and diastolic blood pressure (DBP) from time domain features extracted from auscultatory and oscillometric waveforms and using Gaussian Mixture Models and Hidden Markov Model (GMM-HMM). The nine time domain features selected include the cuff pressure (CP), the cardiac pe...
Conference Paper
Full-text available
Speech based continuous emotion prediction systems have predominantly been based on complex non-linear back-ends, with an increasing attention on long-short term memory recurrent neural networks. While this has led to accurate predictions, complex models may suffer from issues with interpretability, model selection and overfitting. In this paper, w...
Conference Paper
This paper presents a novel method to estimate systolic blood pressure (SBP) and diastolic blood pressure (DBP) from time domain features extracted from auscultatory waveforms (AWs) and using a Gaussian Mixture Models and Hidden Markov Model (GMM-HMM) classification approach. The three time domain features selected include the cuff pressure (CP), t...
Article
Full-text available
The rate of change of frequency (RoCoF) following any disturbance in low voltage islanded microgrids is not only relatively high compared to conventional systems, but also becomes location specific due to the presence of highly resistive lines. This makes the disturbance detection challenging. A novel real‐time disturbance detection technique for l...
Article
Full-text available
The rate of change of frequency (RoCoF) following any disturbance in low voltage islanded microgrids is not only relatively high compared to conventional systems, but also becomes location specific due to the presence of highly resistive lines. This makes the disturbance detection challenging. A novel real-time disturbance detection technique for l...
Article
Full-text available
DC arc faults, especially series arcing, can occur in photovoltaic (PV) systems and pose a challenging detection and protection problem. Machine learning based methods are increasingly being used for fault diagnosis applications. However, the performance of such detection algorithms will degrade because of variations between the source domain data...
Conference Paper
Full-text available
Application of depth-selective filtering to modulated-sparse light fields towards reducing the DSP and memory complexities in real-time light field processing is investigated. A modulated-sparse light field is obtained by spatially windowing an original 4-D light-field signal, which is subsequently processed by 4-D depth-selective filters to achiev...
Preprint
Full-text available
Cepstral normalisation is widely employed in replay detection systems. However, incorporating some information that is lost during normalisation may be useful. Additionally, anti-spoofing systems may further benefit from not treating all speech frames identically. In this paper, we separate speech information based on two different criteria and mod...
Conference Paper
Full-text available
Adaptation of deep neural network (DNN) based language identification models is still a challenging area of research. Recently, state-of-the-art approaches to short duration language identification task have made use of bidirectional long short-term memory (BLSTM) recurrent neural network (RNN) language identification models. Although this enables...
Article
Continuous time-varying prediction of emotions based on speech in terms of attributes (i.e. arousal) has received considerable attention in the past few years. However, the variability introduced by factors not related to emotion, such as speaker and phonetic variability, which in turn may lead to less reliable models and less accurate emotion pred...
Data
ASVSpoof Version 2.0 Speaker Specific file set for the paper titled "Use of Claimed Speaker Models for Replay Detection ". Also available in http://www2.ee.unsw.edu.au/ASVspoof/
Conference Paper
Full-text available
Replay attacks are the simplest form of spoofing attacks on automatic speaker verification (ASV) systems and consequently the detection of these attacks is a critical research problem. Currently, most research on replay detection focuses on developing a stand-alone countermeasure that runs independently of a speaker verification system by training...
Article
High impedance faults (HIFs) on overhead power lines are known to cause fires. They are difficult to detect using conventional protection relays because the fault current is insufficient to cause tripping. The delay in detecting HIFs can result in severe bushfires and energy losses; hence a high throughput, low latency detection scheme needs to be...
Conference Paper
Full-text available
This paper presents a novel framework for speech-based continuous emotion prediction. The proposed model characterises the perceived emotion estimation as time-invariant responses to salient events. Then arousal and valence variation over time is modelded as the ouput of a parallel array of time-invariant filters where each filter represents a sali...
Article
Full-text available
In this paper, we propose a generalized variability model as an extension to the total variability model. While the total variability model employs a standard normal prior distribution in its typical setup, the proposed generalized variability model relaxes this assumption and allows the latent variable distribution to be a mixture of Gaussians. Th...