In line with the UTAR Popular Science Webinar Series, UTAR Division of Community and International Networking (DCInterNet) organised a talk titled “Intel AI Applications: A Closer Look on Consumer and In House Use Cases” via Zoom on 2 July 2021. The webinar saw 239 participants in total.   

The purpose of the webinar was to help UTAR staff, students and the general public to better understand the applications and implications of AI these days. Besides, the participants also hoped to gain knowledge on a few hand-picked applications that Intel approaches in the consumer space as well as in-house applications in Engineering, Human Resources, IT (information technology) and Finance.

Invited to deliver the talk was Intel Malaysia Pre-silicon Validation Engineer Dr Oong Tatt Hee. The webinar was moderated by UTAR Faculty of Information and Communication Technology (FICT) Department of Computer Science Assoc Prof Ts Dr Tan Hung Khoon.

Dr Oong started his talk session with the introduction of AI (Artificial intelligence) and analytics. He introduced the four types of machine learning run by AI, namely supervised learning, unsupervised learning, semi-supervised learning and reinforcement learning.

He explained, “We have seen the impact of AI everywhere. AI refers to the ability of a machine to learn from experience and perform in a manner corresponding to human intelligence and discernment. Supervised learning is the most widely used machine learning in AI as it uses labelled data and teaches model functions. On the other hand, unsupervised learning is the opposite of supervised learning as it extracts the data from inside. So, the data is not labelled because the AI model will learn by itself. Semi-supervised learning is an approach that uses the combination of labelled and unlabelled data to extract data from the inside. Reinforcement learning is a special type of machine learning because this approach could make the best possible decision through trials.”

According to Dr Oong, the data needs to go through a lot of processes before they could generate insights for business, operation or security. First, the data has to go through processes like creating, transmitting, ingesting, integrating, staging, cleaning and normalising. Then, the data will go through various ensemble methods to generate the outcome.

Dr Oong said, “AI is interdisciplinary because it involves many different fields. For example, AI could do image recognition, object detection and image segmentation which are all related to images. It could also do Natural Language Processing (NLP) and speech covert to text or the reverse. It also involves recommender systems as well as mathematics and statistics. Once the data is ready to use, the approach to unlock the valuable insights depends on the data type and the use case requirement. Apart from data analytics, machine learning, deep learning and reinforcement learning, other advanced approaches of AI are comprised of symbolic reasoning, analogical reasoning, evolutionary computing, Bayes methods, etc.”

Dr Oong then moved on to share the different general use cases of AI applications. According to him, the 12 industrial sectors that use AI technology include agriculture, energy, education, government, finance, healthcare, industrial, media, retail, smart home, telecom and transport.”

After showing the general cases of AI applications, Dr Oong gave examples of AI applications to the participants. He said, “In agriculture, AI is mainly used to achieve higher yields and efficiency. For instance, it is used to deploy the amount of water, nutrients or sunlight to a particular crop or plant. In the energy sector, AI is used to predict the real-time intelligence and location of the storage of power as well as reducing the consumption of power. In the education sector, AI application is mainly used to transform learning. It could do cheating detection and provide a virtual mentor or foreign language instruction (translation). The government uses AI technology systems to enhance safety such as preventing crime. The finance sector uses AI to turn data into value, while the healthcare sector uses AI to revolutionise patients’ outcomes.”

As a worker in Intel Malaysia, Dr Oong also highlighted the application of Intel products run by the AI system to strengthen medical technology. Huiyi Huiying Medical Technology (HYHY) is a company that specialises in the application and development of computer vision and deep learning technologies. According to Dr Oong, HYHY applied technologies such as 2nd Gen Intel Xeon Scalable processors with Intel Deep Learning Boost and software optimisation tools like Intel Distribution of OpenVINO toolkits and Intel Distribution for Python to help optimise the performance of the full-cycle AI medical imaging solution.

Besides, Intel is also a trusted partner of GE Healthcare for 15 years. GE Healthcare uses Intel products such as Intel Core processors on Venue to meet the computational needs as well as Intel Distribution of the OpenVINO toolkit to quickly meet the AI applications’ performance requirements during the development of Vivid Ultra Edition features.

An interactive Q&A session was held and it was followed by a closing speech by DCInterNet Deputy Dean Dr Chen, I-Chi.

Dr Chen expressed her gratitude to Dr Oong and said, “We definitely learnt a lot through this talk. In the future, if you have any inquiries about AI technology, please feel free to contact our office. We are happy to introduce UTAR’s AI experts to you. On behalf of UTAR, I would also like to thank you for participating in this webinar. Please stay safe and healthy.”

The webinar ended with a group photo session.