With the rise of the Industrial Revolution 4.0 (IR 4.0), the need for computer vision in many applications has become indispensable. In Malaysia, industries as well as research agencies and universities, have been actively involved in applying and carrying out research in this field. A key component in successful computer vision applications is the ability of computer algorithms in making accurate decisions.

Traditionally, machine learning approaches have achieved good performance. Nevertheless, with the advent of deep learning in the past decade, researchers have increasingly focused on taking advantage of the benefits it provides. Thus, due to its higher performance as well as its adaptability, deep learning has become very popular in computer vision applications. In this Summer School, we are proposing topics that cover both the traditional machine learning (ML) and deep learning (DL) approaches, so that researchers will benefit from the strengths of both, whilst also gaining a historical understanding for the need and importance of the transition from ML to DL in modern computer vision. These important topics and their applications in computer vision will be delivered by prominent national and international speakers. Due to the Covid-19 situation, this Summer School will be conducted in a hybrid fashion, catering for both those who prefer face-to-face participation as well as those attending virtually. All lectures and hands-on practicals will be conducted via live online sessions, while poster sessions will be held in the conference room.

Over the 5 days, we propose 6 lecture/tutorial sessions covering various topics of Machine Learning (ML) and Deep Learning (DL) applications in computer vision, 4 hands-on sessions, and 4 other sessions of forums and discussions with relevant industries. 

The list of topics to be covered during the School is as summarized below:

Technical Topics

1. Introduction to Machine Learning – Classical methods
2. Object Recognition/Classification using Classical Methods
3. Fundamental Concepts in Deep Learning
4. Variations and Advantages of Deep Learning Frameworks
5. Object Recognition/Classification using Deep Learning
6. Transfer Learning and Reinforcement Learning
7. Deep Learning for Video Processing
8. Deep Learning Future and the Way Forward (Forum)

The 5-day hybrid program will cover all aspects of ML and DL, and their applications in computer vision. To gain better understanding as well as make the school more exciting and interactive, hands-on sessions will be conducted during the afternoon time slots. For ML, the school will start with an introduction before covering more advanced topics such object recognition and classification. Similarly, for DL, the school will initially cover the fundamental aspect of DL before proceeding with more advanced topics such as deep convolutional neural network, transfer and deep reinforcement learning, and deep learning architectures. Besides the regular lecture sessions, there will also be online live discussions or forums between the participants and speakers to discuss particular topics of interest in computer vision. There will also be poster sessions for participants to showcase their current work. At the end of the school, participants will be exposed to the fundamental and advanced knowledge of DL and big data, and their applications in computer vision. The expected outcome of this Summer School would be that participants would be able to create better solutions and explore greater perspectives in their domains of interest within this growing field of research.