Research into data science and artificial intelligence is non-stop. As the world speeds up or slows down in other ways, there are always researchers looking to push the boundaries of what we can do with data, automation, and robotics. In the APAC region, including parts of Asia and the Pacific region (especially in Australia), there are plenty of research institutions that have made some interesting developments in recent years and we’d like to shine a spotlight on some APAC AI Research Universities.
The University of Tokyo: Multiple Labs
Within the University of Tokyo, there are a number of exciting research labs, each with its own unique focus has part of this list of APAC AI research universities.
The Sato Lab, as part of the Department of Complexity Science and Engineering, focus on using machine learning to tackle real-world problems, as opposed to “researching for research’s sake.” Lead by Issei Sato, this lab is working on projects like developing a supporting system for a medical image analysis collaborating with the University of Tokyo hospital.
With some similarities, the Sugiyama-Yokoya-Ishida Lab and each branch, led by a different professor, have their own unique goals. The Sugiyama lab focuses on the theme of “how intelligent can computers be?” and looks into how you can apply machine learning in real-world scenarios. The Yokoya lab dives into computational imaging from space to create map information and plans for the future. Lastly, the Ishida Lab focuses on developing better machine learning algorithms to, in turn, develop better and more practical machine learning technology.
The Miyao Group, lead by Yusuke Miyao, heavily focuses on the mathematical models and algorithms behind NLP, such as syntactic parsing, semantic parsing, knowledge discovery, question answering, and more. Recently, they build the NIILC Question Answering Dataset, which scours Wikipedia for answers to questions.
As another lab that focuses on NLP, the Aizawa Lab researches subjects in text and media studies such as text mining, machine reading comprehension, and human language activities. They provide a lot of their research openly, including one of their main topics in “establishing a common understanding between humans and computers through natural language texts.”
Similar to the University of Tokyo, and some other APAC AI research universities, there are a number of interesting AI labs within the Kyoto University Artificial Intelligence Research Unit. This includes the Shimodaira Lab (statistics and machine learning), the Yoshikawa & Ma Laboratory (big data and data mining), the Kashima Machine Learning and Data Mining Research Laboratory (data analysis technologies based on statistical machine learning), and the Kurohashi-Chu-Murawaki Lab (NLP and NLU).
Tokyo Institute of Technology: Okazaki Laboratory
Focusing on NLP, the Okazaki Laboratory researches for both theory and applications. The lab studies all things NLP, such as sentiment analysis, machine translation, natural language generation, and natural language understanding. Their latest research focused on Multimodal Pretraining and BERT.
Seoul National University
The Machine Intelligence Lab
The Machine Intelligence Lab mainly researches NLP, deep learning, and their applications. They stand out with their collaborative approach in working with other academic institutions as well as research wings of industry leaders, like Microsoft, Samsung, Adobe, and more.
Intelligent Data System Laboratory
As a lab that dates back to 1992 on this list of APAC AI research universities, the Intelligent Data System Laboratory focuses on data mining, data processing, information retrieval, recommendation, natural language processing, and deep learning. They’ve recently been exploring BERT to improve sentence representations in a self-supervised setting.
Also featuring a unique assortment of individual labs, the Graduate School of AI at Korea Advanced Institute of Science and Technology (KAIST GSAI), aka just KAIST AI, is one of the highest-ranked AI research institutes in the world. With research that covers nearly every topic under the AI umbrella, their research is not just theoretical but has found applications with tech companies including Google Brain, IBM Watson lab, Disney Research, NVIDIA, and so on. The lab is making news constantly, such as their work on developing a new chip for efficient deep reinforcement learning, working with K-pop groups on creating a metaverse, and more.
University of Sydney: UBTECH Sydney Artificial Intelligence Centre
The Uintah Basin Technical College as part of the University of Sydney is committed to advancing AI to endow machines with the capabilities of perceiving, learning, reasoning, and behavior. They have videos of some of their AI projects in action, such as this one on multi-object tracking.
University of Technology Sydney: AAII (The Australian Artificial Intelligence Institute)
Offering a broad range of research topics across a number of different research labs internally, the AAII focuses on everything from theoretical machine learning research to intelligent drones. As a newer lab, they’re hitting the ground running with research into brain-computer interface applications, representation learning on large scale healthcare administrative data, deep learning smart systems for educational management, and more.
University of New South Wales: D2 Lab (Data Dynamics Lab)
The D2 Lab focuses on a little bit of everything related to explainable and generalizable data, such as efficient data mining, ML/DL, systems and interfaces, and human-computer interaction. They offer a lot of demos that show their research in action, such as this series on their work on the Internet of Things.
NUS (National University of Singapore) – Natural Language Processing Group
As the name suggests, the NUS Natural Language Processing Group exclusively studies NLP. Lead by Professor NG Hwee Tou, this group offers open-source software, data, and insights into their publications. As paraphrasing is common, they’ve developed PEM, which is an automatic metric designed to evaluate the quality of paraphrases.
SUTD (Singapore University of Technology and Design) – StatNLP
Also devoted to NLP, the StatNLP lab offers a lot of resources and datasets for anyone interested. For example, the Multilingual Geoquery, which is a multilingual dataset for Geoquery. Each instance is a sentence annotated with its meaning representations.
Conclusion of APAC AI Research Universities
This list isn’t exhaustive of course. There are plenty of other research institutions, universities, and companies with research units that are doing amazing work. Who would you add to the list? We’re open to updating this with any place that we may have missed.
Learn more about these APAC AI Research Universities at ODSC APAC 2021
ODSC APAC 2021 will feature a number of focus areas devoted to hot topics in data science and artificial intelligence, such as machine and deep learning, MLOps, and research frontiers. A number of the research institutions listed above will be well-represented by both speakers and attendees, so if you’re interested in the cutting edge of AI, then you’ll definitely want to check out the event. Register here for 40% off all ticket types and attend the virtual event this September 15th-16th.