Meet Shin Fang Ch’ng, a robotic vision PhD student with her sights set on leadership

Story written by Dr Sarah Keenihan, AIML

When she looks forward ten years, Shin Fang Ch’ng imagines herself leading a research team in one of the world’s big technology companies. 

“You know, Facebook, Tesla, Microsoft, somewhere like that,” she says. 

“That would be my dream job, to be able to convert research into tangible products that people find useful.” 

Shin Fang is a PhD candidate at the Australian Institute for Machine Learning (AIML), University of Adelaide. Here, she’s developing the research and leadership skills in preparation for the career she has in mind. 

“In my PhD, I’m working on robotic perception,” she explains. 

“My goal is to allow a robot, which is just a kind of specialised machine, to be able to operate and navigate in the real world, just like humans do.” 

Robots that ‘see’ and ‘understand’

The field of robotic perception aims to give machines the ability to make decisions depending on where they are. For example, a self-driving car has to be able to know its location in relation to traffic lights, other vehicles and pedestrians, and navigate to a destination without hitting any obstacles. 

Through her research, Shin Fang aims to improve how robotic vision systems perform a task known as SLAM – Simultaneous Localisation and Mapping. 

“SLAM estimates the robot's motion and the 3D geometry of the environment by transforming pixel values seen through a camera,” Shin Fang says. 

“But SLAM has to be both fast and able to cope if the information isn’t perfect – and it’s these aspects I’m working on in particular.” 

Shin Fang uses two different approaches to test how well her algorithms accurately recreate real scenes.

“We take a good look at each new model, and make an overall assessment of how similar it is to the original environment,” she says. 

This is a qualitative assessment, based on observation. The other approach is quantitative, and involves data. 

“For this one, we use measurements from the original scene to calculate the errors introduced when we used the algorithms to recreate it,” says Shin Fang. 

Developing new skills

Shin Fang only has a couple of months left before she expects her PhD will be completed. She has published research papers and presented them at computer vision and robotic conferences, and will submit a thesis.   

After that, she’s deciding whether to move into an industry role, or stay in academia for a while longer. 

“I really like research; I find it so exciting and challenging,” she says. 

“So I’d have to say I’m leaning in the research direction for the moment.” 

In addition to her research activities, Shin Fang was recently engaged as a tutor at the Adelaide node of the 2021 Robotic Vision Summer School. 

“I was responsible for communicating with the Summer School organisers and tutors from other nodes, and organising the activities in the Adelaide node to ensure the program went smoothly,” she says. 

“I also provided advice to students when they participated in some of their coding and workshop tasks.”

“I think these skills are really useful in a career sense,” Shin Fang says. 

Take a chance

Prior to starting her PhD in 2017, Shin Fang completed an Electronic Engineering degree with honours, in Kuala Lumpur. 

While working as an engineer, she had reached out to her now PhD supervisor Professor Tat-Jun Chin.    

“I gained a scholarship to come to AIML, and here I am today,” Shin Fang explains. 

“I feel very grateful to have this opportunity, to be able to do my PhD in Australia.”  

Shin Fang’s research at AIML is conducted in collaboration with the Australian Centre for Robotic Vision .