Computational Cognition Vision and Learning (CCVL) Postdoctoral Fellow

Thursday 18th August 2022

Organization Johns Hopkins University
Location Baltimore, Maryland
Title Computational Cognition Vision and Learning (CCVL) Postdoctoral Fellow
Email Address

The goals of CCVL (Computational Cognition Vision and Learning) are to develop mathematical and computational AI models of vision and its relations to other cognitive abilities including language. These models should ideally have the same abilities as humans including the capability of building models of the 3D world. CCVL consists of 20-30 graduate students and postdoctoral researchers. It is part of the AI Vision Collective at JHU (members include Rama Chellappa and Vishal Patel), the Center for Language and Speech Processing (CLSP), and MINDS.

CCVL runs a summer internship program that involves 10-20 undergraduate interns. Postdoctoral researchers are encouraged to work with other research group members. CCVL has a strong record in publications at major conferences including CVPR, ICCV, ECCV, ICLR and NeurIPS with 12 papers at CVPR 2022 and 10 papers at ECCV 2022.

1. Candidates will have a PhD degree and publication record in a statistical science, machine learning, or other data analysis field.
2. An ideal candidate would have knowledge of CNNs, transformers, unsupervised learning, geometry, Bayesian techniques, computer graphics rendering models, and adversarial examiners.
3. Candidates with interests in analysis by synthesis and the need to test models on out-of-distribution data are encouraged.
4. A strong publication record in major conferences is strongly desirable.
5. Candidates should have strong motivations for research and desire to address challenging problems, be a team player and willing to work with graduate students, help write grant proposals, and help manage the research group.

Positions are available now with an initial appointment for one year but with the possibility of extension.

If interested, contact Alan Yuille at enclosing CV, statement of interest, list of selected publications, and names of people who can give recommendations.