Please see main research areas of CGM Lab:
Our vision is to transform the way in which archaeology research is being performed. Oftentimes, archaeologists not only explore the past, but also do it in manual, Sisyphean ways that should belong to the past. Instead, imagine a world where an artifact is automatically drawn by a computer, rather than by an archaeological artist; where searching for corollaries is performed by a Google-like system; where worn-out artifacts are ”brought-back” to life. Interestingly, these tasks are classical computer vision tasks. Archaeological artifacts, however, are much harder to deal with than usual, since they are broken and eroded after laying underground for thousands of years. We work on all facets of this field.
When looking at the world, people do not necessarily perceive the real physical measurable properties. Can computers mimic what people perceive, rather than what they see? Within this topic we explore saliency detection in 2D and in 3D, memorability prediction, and visual illusions.
The aim is to develop algorithms, learning models and technologies that ”understand” 3D data from its geometric properties. Analyzing 3D visual data is vital in many aspects of our lives, such as design, entertainment, scientific visualization, autonomous cars, robotics etc. We work on numerous facets of this field, among which are classification, segmentation, similarity, detection, shape completion, edge detection, surface saliency and visibility detection.
We have done extensive work on recognition in images and videos, including multi-label image classification and event detection in video.
Haptics is to touch what Optics is to vision. Our ultimate goal is to digitize the sense of touch, to enable multi-modal immersive experiences involving multiple senses, including touch.
We have done extensive work on the search, design, and training of Neural Architectures.
