Full Time Postdocs in Machine Learning
Description:
Two Postdoctoral Researchers in Innovations in Formalized Learning for Open Worlds
Dr. Boult and his team at UCCS are seeking highly motivated postdoctoral researchers with a strong record of accomplishment in machine learning and computer vision, ideally with incremental learning, deep learning, or open-set learning. Classic machine learning and AI make a closed-world assumption, leading to fragile systems that cannot label novel inputs as unknown nor handle novel variations in the world. Very few existing systems can detect novelty and then learn from previous unknowns – almost all systems misclassify the unknown. At the core of intelligence is the ability to recognize when we do not know something, analyze the need to learn about it, and then, when needed, to adapt and learn it – a process we call open-world learning. We seek a candidate to join our multi-institution efforts exploring four main research directions: 1) Formalized models for open world learning with extreme-value theory, 2) Semi-supervised and unsupervised incremental learning 3) Open world learning for object classification and detection, and 4) Open world learning in visual observation of games. The PostDocs are expected to both do research, develop software, write papers, and lead/works with students.
Five or more years of Computer vision and/or Machine Learning research and development experience required. Developments will be in Python and C++ on Linux and significant systems/software engineering experience in one or both languages preferred. A Ph.D. is preferred but an MS with significant experience will be considered. International candidates with strong English skills will be considered; we can manage the visa issues. The University of Colorado is an equal opportunity employer; women and underrepresented populations are encouraged to apply.
If you are interested, please apply at
Contact: Terry Boult (tboult@vast.uccs.edu)
