HO-PHUOC Tien
Courses
- Image processing, Visual perception
- Signal processing, Random signal processing, Spectral analysis
Research Interests
My interests focus on image/video processing including the following topics.
Visual attention / Saliency model: When looking at a scene human eyes concentrate on some specific or salient regions in order to get useful visual information. The ability of instantaneous focus on visually important regions is called visual attention. While a machine supposed to possess such ability may lead to many interesting applications, a great interest has been put on how to understand the mechanism of visual attention and how to predict salient regions where persons are likely to look. The latter task can be carried out by a saliency model or visual attention model. This model can certainly help to improve the on-going machine vision, as well as have promising applications such as compression and advertising.
Image superposed with human fixations (left) and the corresponding human fixation map (right)
Recognition: This problem has been showing a great interest in human-computer interaction and industrial applications. A broad spectrum of recognition topics include human characteristics, e.g. face, fingerprint, iris, and activity, and non-human recognition such as object or fish trajectory.
Super-resolution and Demosaicing: This topic of image processing is quite important for image sensors. On one hand, demosaicing is used to reconstruct a full color image from a mosaic one in which each pixel has only one color, i.e. missing two other colors. On the other hand, super-resolution is required to enlarge the size of an image while ensuring its visual quality. These two problems can be treated together or separately. An expectation is to devise an efficient method to be implementable on image sensors.
Mosaic image (left) and its reconstructed full color image (right)