Matthew Kowal

Matthew Kowal

PhD Student - Deep Learning for Video Understanding

About Me

I am currently doing my PhD at York University under the supervision of Dr. Kostantinos Derpanis. I am a postgraduate affiliate at the Vector Institute and the Lead Scientist in Residence at NextAI. My research interests are in the domain of explainable computer vision with a focus on image and video understanding tasks.

I completed a Bachelors of Applied Science (B.A.Sc) in Applied Mathematics and Engineering with a specialization in Mechanical Engineering at Queen’s University in Kingston, Ontario, Canada. After graduating from my Bachelors degree, I worked at Morrison Hershfield as a mechanical design engineer (in training). I designed buildings, labratories, and condos, with a team of other mechanical, evironmental, electrical, and control engineers. I then obtained my M.Sc at the Ryerson Vision Lab in August 2020 under the co-supervision of Dr. Neil Bruce and Dr. Kosta Derpanis. My M.Sc thesis focused on multi-modal action recognition.

As the Lead Scientist in Residence (SiR) at NextAI, I act as a technical consultant for multiple AI-based startups in Toronto. Some of the companies I have worked with are Origami-XR, Future Fertility, VideoLogic, NoLeak Defence, Argentum.

My hobbies include health and fitness, competitive Super Smash Bros. Melee, birds, close up magic, and progressive house music.

News

  • I gave a talk at Vector’s Endless Summer School program on Current Trends in Computer Vision and a CVPR 2022 Recap
  • Paper accepted to the International Journal of Computer Vision (IJCV) - SegMix: Co-occurrence Driven Mixup for Semantic Segmentation and Adversarial Robustness. Paper.
  • New Preprint available on Arxiv - Quantifying and Learning Static vs. Dynamic Information in Deep Spatiotemporal Networks. Paper
  • I presented a spolight presentation at the Explainable AI for Computer Vision Workshop at CVPR 2022. You can watch the recorded talk here.
  • Paper Accpted to CVPR 2022 - A Deeper Dive Into What Deep Spatiotemporal Networks Encode: Quantifying Static vs. Dynamic Information. Paper and Project Page.
  • Paper Accepted to ICCV 2021 - Global Pooling, More than Meets the Eye: Position Information is Encoded Channel-Wise in CNNs. Paper.
  • Paper Accepted to BMVC 2021 - Simpler Does It: Generating Semantic Labels with Objectness Guidance. Paper.
  • Paper Accepted to ICLR 2021 - Shape or Texture: Understanding Discriminative Features in CNNs. Paper.
  • Paper Accepted as an Oral to BMVC 2020 - Feature Binding with Category-Dependant MixUp for Semantic Segmentation and Adversarial Robustness. Paper.