Here you can read about all things related to my research. Hopefully some of them are useful to your own.
Mandatory social profiles here:
Google Scholar profile: https://scholar.google.ca/citations?user=3haLp1AAAAAJ&hl=en
Reading Group Slides
We have a weekly reading group. And once or twice a semester, I go into a paper into great depth and make some slides. This takes a surprising amount of time. Often I publish them here in the hope that these days analyzing these great works are useful to others.
I’ve tried to note within the slides those parts that I was confused about.
Generative Adversarial Networks
Two neural networks, competing with each. That’s what Adversarial Networks are. We can use this approach to generate novel images!
Here is a writeup with slides that cover generative adversarial networks:
The slides cover Goodfellow et al., and Radford et al. (mostly focused on Radford et al.)
Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Bengio, Y. (2014). Generative Adversarial Nets. NIPS, PDF
Radford, A., Metz, L., & Chintala, S. (2016). Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks. In ICLR. arxiv
Classify and retrieve skin diseased images
Predict neurodevelopmental outcomes in brain scans
We worked on convolutional neural networks for brain connectomes to predict neurodevelopmental outcome of preterm infants.
Predict MS disability and segment spinal cords
My master’s research was devoted to segmenting the spinal cord in MRI volumes, and predict the disability level of multiple sclerosis patients from features in their spinal cord.
Detect kidney tumors in ultrasound videos
Sometimes, people ask me to talk about my research. Here are the slides for some of the talks. They also serve as a nice high-level overview of what I research.