RMIT University, February 2019 - May 2023
I completed my Doctor of Philosophy in Computer Science with an RTP Fee offset Scholarship, under the supervision of Prof. Flora Salim, Dr. Piotr Koniusz, and Dr. Wei Shao. My thesis is entitled Spatio-temporal Deep Learning. Check Publications for more details on my research.
During my doctorate, I was a privileged recipient of Data61/CSIRO PhD stipends and top-up scholarships, valued at $ 27,596 and $10,000 p.a. respectively.
I am a part of CRUISE (Context Recognition, Urban Sensing, and Intelligence) research group led by Dr. Flora Salim.
Deakin University, 2017
Thesis: A feasibility study of automated detection of courtship display in male guppies [slides]
I combined multiple techniques, including machine vision, Binary Large Object analysis, Discrete Fourier Transform, and Discriminant Function Analysis, and develop novel ones, to make an algorithm to classify fish behaviour. My algorithm achieved an accuracy of 94.2%, Diagnostic Odds Ratio of 76.3, and p-value of 0.0026 using Barnard’s test.
Extensive use of programming in MATLAB
First degree honours
Supervisor: Professor John Endler
The University of Melbourne, July 2012
Learning Areas: Physics and Biology
Placements in Scotch College and Strathmore Secondary College
The University of Melbourne, November 2010
Majors: Physics and Physiology with specialisation in Neuroscience
Exchange and Study Abroad Program, UCLA and UCSD, 2009
I was selected to participate in the competitive Melbourne Global Mobility Exchange and Study Abroad Program where I extended my diversity of international educational experience. Exchange Program with University of California, San Diego. Study Abroad Program with University of California, Los Angeles.
MGM scholarship, 2009
I was awarded $ 2,500 MGM (Melbourne Global Mobility) Scholarship as a part of an academic-merit based scholarship for exchange students.
Research Assistant, Melbourne Neuropsychiatry Centre, Semester 2, 2010
Applied graph theory on white matter connectivity from fMRI data of Friederich’s Ataxia patients.
I extracted the connectome from fMRI data of Friederich’s Ataxia patients and control group to model white matter connectivity. I compared numerous graph theory metrices, such as average path length, between the brain of healthy and Friederich’s Ataxia patients. Programmed in MATLAB
Figure taken from: Zalesky A, Akhlaghi H, Corben LA, Bradshaw JL, Delatycki MB, Storey E, Georgiou-Karistianis N, Egan GF. Cerebello-cerebral connectivity deficits in Friedreich ataxia. Brain Struct Funct. 2014 May;219(3):969-81. doi: 10.1007/s00429-013-0547-1. Epub 2013 Apr 7. PMID: 23563750.
Undergraduate Research Project, Semester 2, 2010
I applied Fourier transform on 7 vowels common in English language, effectively forming a spectrogram. I extracted a spectral template (temporal slice from the spectrograms, 30 milliseconds after the onset) to model long term memory template and performed various neurolgically inspired processing on the spectral template. This model performed with 100% accuracy among the 7 vowels.
Programmed in Python