Machine learning and computer vision for underwater image enhancement

External Member

Mehrdad Shoeiby

Description

Abstract: In this project, students have the opportunity to work a unique set of data, that is,  stereo RGB and multi-spectral images from the Great Barrier Reef, as well as local water parameters that can help with the image restoration process. The objective is to remove the effect of the water from coral images as if the images were taken in air. The students will have access to GPU clusters and interact with domain experts, as well as work on a project with potentially significant environmental benefit.

Contact: Mehrdad Shoeiby

 

Updated:  1 June 2019/Responsible Officer:  Dean, CECS/Page Contact:  CECS Marketing