_{Hi there! I’m Chai Ai Fang, a PhD candidate at Monash University Malaysia, diving deep into the fascinating world of AI, particularly in the realm of human digital representation. My research focuses on generating natural and realistic human talking face videos using audio and image inputs—essentially, bringing virtual humans to life! But it wasn’t always like this. Let me take you through my journey, which has taken me from Monash University to Campus Fryslân, as I explore the cutting-edge field of speech technology.}

Turning Passion into Research: Realistic Talking Faces and Hate Speech Mitigation

Before my PhD journey, I completed my Bachelor’s degree in Computer Science at Universiti Sains Malaysia, specializing in Artificial Intelligence. This was where I first stumbled upon my passion for AI. As I explored various AI models, I was especially drawn to computer vision and speech synthesis—fields that, as it turned out, would play a significant role in my current research.

Now, as a PhD candidate, I’m investigating how we can create more accurate, realistic talking face videos using AI. It’s an exciting challenge! My research focuses on how advanced AI technologies can help generate realistic and natural human talking face videos while also considering the ethical aspects involved.

A significant part of my work involves mitigating hate speech in audio-based content. I’m exploring how large language models (LLMs) and speech synthesis systems can serve as pre-processing tools to filter out potentially harmful content from generated videos. This ensures that the videos we create are not only realistic but also ethically sound—something I believe is essential as we move forward with this technology.

In 2024, my home campus launched the PhD Global Mobility Program, an incredible opportunity for me to explore research environments at other universities. This program allowed me to broaden my academic horizons and engage with experts in my field, aligning perfectly with my research interests.

❝Now, as a PhD candidate, I’m investigating how we can create more accurate, realistic talking face videos using AI.❞

Campus Fryslân: A Researcher's Dream

My co-supervisor has a senior colleague at Campus Fryslân, who is an expert in Speech Technology. Given my keen interest in speech synthesis, speech recognition, and emotion classification in speech—all closely related to my research area—Campus Fryslân became a top choice for me. Additionally, I applied to other universities and received offers from both the University of Science and Technology Beijing (USTB) and Campus Fryslân. Being Asian, I was particularly eager to experience Western research culture firsthand, and Campus Fryslån’s high academic ranking further reinforced my decision.

The host supervisor at Campus Fryslân was incredibly responsible and welcoming. They went above and beyond by introducing me to various speech technology-related activities, such as the 3rd Dutch Speech Tech Day, which significantly broadened my knowledge and expanded my professional network.

The campus environment was exceptionally comfortable, with well-equipped workspaces and coffee machines readily available throughout. I particularly enjoyed grabbing a coffee during midday—a small but much-appreciated perk! Beyond the excellent facilities, the researchers at Campus Fryslân were incredibly approachable and open to discussing their work and sharing technical insights.

Throughout this journey, I gained valuable knowledge and insights into speech technology, making this experience truly enriching. The opportunity to immerse myself in the Campus Fryslân research culture and learn from experts has been invaluable to my academic and professional growth.

❝Given my keen interest in speech synthesis, speech recognition, and emotion classification in speech—all closely related to my research area—Campus Fryslân became a top choice for me.❞

In Case You’re Still Wondering: What Is Speech Tech and What Can It Do for Society?

1. Enhancing Accessibilty for Invididuals with Disabilities

Speech technology significantly improves accessibility for individuals with disabilities. For people with visual impairments or motor difficulties, voice commands and speech-to-text systems offer an alternative way to interact with technology, enabling them to use smartphones, computers, and smart devices effortlessly.

2. Bridging Language Barriers for Travelers

Speech technology is beneficial not only for individuals with disabilities but also for people traveling to foreign countries who may face language barriers. Travelers who are unfamiliar with the local language and may struggle with typing messages can use speech-to-text features and convert the text into the local language, making communication much easier and more convenient.

3. Streamlining Daily Life with Voice Assistants

Voice assistants like Siri, Google Assistant, and Alexa have become integral parts of daily life, helping users with tasks such as setting reminders, searching for information, and controlling smart home devices. Speech technology enhances communication by providing hands-free and efficient ways to access information and perform various tasks, improving both convenience and productivity.

Looking ahead, I’m excited to advance my research by generating realistic and natural talking face videos that are free from harmful content while preserving genuine human expressions. This balance between authenticity and ethical integrity is at the heart of my work, and I can’t wait to see where this journey takes me!

Join us in shaping how humanity interacts with AI!

- Sign up for our monhtly newsletter

- Visit our Speech Technology MSc programme webpage
- Already convinced? Apply now!