Michael S. Brown

Professor, EECS Department
Member, Centre for Vision Research
Member, IC@L
Faculty Affiliate, Vector Institute
Senior Research Director Samsung Research – Toronto (Part-time)

Website | Email

2021 – 2022 Research Highlights

Developing Methods for Improving Camera Imaging

My research investigates image formation models that explain how 2D images are created from light irradiance incident on a camera’s sensor under different imaging scenarios. My work also examines how such models can be used for different computer vision applications.  Last year, much of my work focused on novel modifications to the camera hardware to improve post-capture understanding of images.  My research lab also examined recent dual-pixel sensor designs that provide the functionality of a rudimentary light-field. This light-field data is now available on virtually call cameras but is ignored.  We have shown that dual-pixel information can be used for many tasks, including reflection removal, depth estimation, and extended depth of field.

Research Highlights

1. Punnappurath A. and Brown M. S. (2020) “Camera ISP Modification to Enable Image De-rendering”, 28^th IS&T Color Imaging Conference, (CIC28), Nov. 2020
2. Le H., Afifi M., and Brown M. S. (2020) “Improving Color Space Conversion for Camera-Captured Images via Wide-Gamut Metadata”, 28^th IS&T Color Imaging Conference, (CIC28), Nov. 2020
3. Abuolaim A., Punnappurath A., Brown M. S. (2020) “Defocus Deblurring Using Dual-Pixel Data”, European Conference on Computer Vision (ECCV’20), Aug. 2020
4. Afifi M. and Brown M.S. (2020) “Deep White-Balance Editing”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR’20), June 2020
5. Punnappurath A., Abuolaim A., Afifi M., Brown M. S. (2020) “Modeling Defocus-Disparity in Dual-Pixel Sensors”, International Conference on Computational Photography (ICCP’20), April 2020