my human-centred visual learning research is to develop vision-based algorithms that aim to make systems usable and useful by focusing on humans, especially their needs and requirements. In particular, the understanding human eye gaze is fundamental for meaningful interpretation of human action and behaviour. With the advent of deep learning and high-resolution cameras, appearance-based gaze estimation has been the centre of the field. However, unlike the other body pose estimation research, eye gaze is very difficult to collect large-scaled natural eye movement dataset, and it is the biggest hurdle to go in the wild despite its strong link to real applications such as augmented reality human-computer interaction, and driving scenarios. In this talk, I will present my recent studies on a large data set generation framework for appearance-based gaze estimation and ongoing human gaze-based attention reading research.