The ECU Computer Vision Lab led by Dr. David Hart uses modern machine learning and artificial intelligence techniques to expand the field of computer vision. Currently, our lab is conducting research in the following fields.
Style transfer involves the transformation of images to adopt the artistic style of a reference photo. At our lab, we explore innovative techniques and algorithms to enable intricate guiding of the style transfer process. By analyzing segmentations, feature statistics, and other properties, we are creating new approaches to give users more control over the creative process.
We actively explore how to expand modern 2D neural network designs into the realm of 3D objects. For 360° photos, 3D models, depth images, or radiance fields, we look to find new approaches to understanding and editing this data.
The ECU Computer Vision Lab works with many collaborators in the medical field. We aim to apply modern computer vision and machine learning techniques to revolutionize healthcare at an individual level. Current collaborations include projects such as home healthcare monitoring through depth sensors (led by Dr. Kamran Sartipi) and sperm tracking for infertility analysis (led by Dr. Cameron Schmidt).
Graph Neural Networks continue to play a major role in many fields of computer vision. As more and more data becomes multimodal, combining images with text and other domains, we believe that graphs will provide a powerful mechanism to increase the capabilities of machine learning techniques.
Taken from https://neo4j.com/developer-blog/demystifying-graph-neural-networks/