3D Vision: Machine Learning Algorithms to Mimic Human Perception
UMass Amherst computer scientist Evangelos Kalogerakis and PhD student Zhaoliang Lun (College of Information and Computer Sciences)–along with Alla Sheffer from the University of British Columbia–have developed a computer program that compares the style similarity of 3D objects. An algorithmically computed style similarity measure that mimics human perception can benefit several computer graphics applications. The program can help designers to create rooms with stylistically compatible furniture (see image above) or organize a 3D model database according to architectural styles. Kalogerakis is an expert in the development of computational techniques that analyze and synthesize visual content, focusing on machine learning algorithms that help people to create 3D models. Learn more.
Image credit: Kalogerakis Research Group