Title:  Deep Learning for Omnidirectional Video Analysis

Faculty: Aykut Erdem (aerdem@ku.edu.tr)

Description:
The Koç University İşbank Artificial Intelligence Center (KUIS AI Center) is looking for a talented, highly motivated postdoctoral researcher to join a project in the emerging field of deep learning for omnidirectional videos. Omnidirectional videos, a.k.a. 360-degree videos have recently gained interest as they provide immersive experiences for users to explore virtual reality environments. Dealing with this kind of spherical visual data, however, requires radically new types of neural architectures and fresh ideas to effectively train these networks. As part of the research team, responsibilities for this position include:

– Developing computationally efficient models to assess the perceptual quality of omnidirectional videos

– Developing visual saliency approaches to predict where humans look at omnidirectional videos

– Providing guidance to PhD and MSc students

Required Qualifications:

– PhD in Computer Science, Electrical Engineering, or a related technical discipline

– Strong background in computer vision and/or machine learning

– Experience with Deep Learning frameworks such as PyTorch, and/or TensorFlow

This is a full-time position that will have a yearly appointment with annual renewal expected for up to three years in total. The start date for the position is flexible, but the ideal candidate would begin in July 2021.