Jiashi Feng is here on a short stay, visiting Vision.

Jiashi received his bachelor's degree from the University of Science and Technology in China. He is currently a PhD student at the National University of Singapore, with a focus on computer vision, machine learning research, and its applications to multimedia. His research is on object classification - the ability for a machine to understand the content of images. He works mostly on feature learning. Features are low-level descriptions of visual objects, such as their shape, color, or texture. The task is to recognize these features  and extract meaningful information from them. Jiashi works with robust structure mining to do this.

He has worked with the Vision and Machine Learning (VML) group at the National University of Singapore on applications of these techniques such as a music recommender, which recognizes the type of event occurring in a video and recommends appropriate background music, and a clothing recommender, which takes into account both what you're wearing and what kind of situations you will be facing during the day in order to recommend clothes that would complement both. In addition to this, he's also generally interested in machine learning and optimization, particularly robust and online optimization.

At ICSI, he works with Vision on objection classification through the use of attributes, which, in contrast to features, are mid-level descriptions of objects. For example, while features of a car might be that it is red and smooth, an attribute is that it has the shape and texture of a car. He will work on ways that computers can automatically discover attributes in images.

Jiahi enjoys running, watching movies, listening to music, and reading books.