About Me
I am a first-year Doctoral student as York University under the supervision of Dr. Marcus Brubaker and Dr. Kosta Derpanis. I am interested in computer vision and machine learning. More specifically, I am interested in working on methods that can leverage large amounts of unlabeled data. This may include elements of transfer learning, generative modeling, and/or the formulation of rich inductive priors.
News
- Mar 31, 2021 Accepted invitation to the Vector Institute Postgraduate Affiliate Program
- Sep 25, 2020 Paper accepted to Neurips 2020: Wavelet Flow: Fast Training of High Resolution Normalizing Flows
- Aug 6, 2020 Successfully defended Master's thesis
- Apr 29, 2020 Accepted VISTA Doctoral scholarship at York University
- Mar 11, 2020 Ended internship at Borealis AI
- Mar 4, 2020 Accepted Offer of Admission to York University Computer Science PhD. Program
- Nov 12, 2019 Started internship at Borealis AI
- Oct 30, 2019 Accepted MITACS Accelerate award in partnership with Borealis AI
- May 2, 2018 Accepted VISTA Masters scholarship at York University
- Apr 1, 2018 Accepted the Ontario Graduate Scholarship (OGS)
Projects
Wavelet Flow: Fast Training of High Resolution Normalizing Flows
Sep 25, 2020
Normalizing flows are a class of probabilistic generative models which allow for both fast density computation and efficient sampling and are effective at modelling complex distributions like images. A drawback among current methods is their significant training cost, sometimes requiring months of GPU training time toachieve state-of-the-art results. This paper introduces Wavelet Flow, a multi-scale, normalizing flow architecture based on wavelets. A Wavelet Flow has an explicit representation of signal scale that inherently includes models of lower resolution signals and conditional generation of higher resolution signals, i.e., super resolution. A major advantage of Wavelet Flow is the ability to construct generative models for high resolution data (e.g., 1024×1024 images) that are impractical with previous models. Furthermore, Wavelet Flow is competitive with previous normalizing flows in terms of bits per dimension on standard (low resolution) benchmarks while being up to 15× faster to train.
Back to Basics: Unsupervised Learning of Optical Flow via Brightness Constancy and Motion Smoothness
Aug 29, 2016

Recently, convolutional networks (convnets) have proven useful for predicting optical flow. Much of this success is predicated on the availability of large datasets that require expensive and involved data acquisition and laborious labeling. To bypass these challenges, we propose an unsupervised approach (i.e., without leveraging groundtruth flow) to train a convnet end-to-end for predicting optical flow between two images. We use a loss function that combines a data term that measures photometric constancy over time with a spatial term that models the expected variation of flow across the image. Together these losses form a proxy measure for losses based on the groundtruth flow. Empirically, we show that a strong convnet baseline trained with the proposed unsupervised approach outperforms the same network trained with supervision on the KITTI dataset.