About me
I started as an assistant professor (tenure-track) at University of Victoria, Canada in 2022, teaching and doing research on deep learning. My research seeks to discover how deep learning works, experimentally discovering insights that improve our understanding and lead to practical improvements that advance the field.
Education
I completed my PhD in Computer Science from the University of Auckland (New Zealand) and was nominated for the Dean’s List award (top 5% of the year). I also hold an MS in Applied Mathematics and graduated with a Distinction Award for my Master’s Thesis (top two in the department). Along the way also interned at Apple doing natural language processing.
Recent Publications
You can find the full list of my work on deep learning here: Google scholar.
- Nondeterminism and Instability in Neural Network Optimization. Cecilia Summers and Michael J. Dinneen. Spotlight, ICML 2021.
- Four Things Everyone Should Know to Improve Batch Normalization. Cecilia Summers and Michael J. Dinneen. ICLR 2020.
- Improved Mixed-Example Data Augmentation. Cecilia Summers and Michael J. Dinneen. WACV 2019.
During my MS, I also published papers on applied mathematics in journals such as Pattern Recognition Letters.
Awards
- 2022-2027 The Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grants
- 2022 The Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Launch Supplements
- 2020 Google scholarship nominee
- 2019 and 2020 PhD Award from Department of Computer Science, University of Auckland
Invited talks
- IBM New York office, 2022.
- Toyota Research Institute, 2022.
- Purdue University, 2021.
- University of California Davis, 2021.
- Queen’s University Belfast, 2021.
- Meta headquarters (aka Facebook), 2018.
Contact
E-mail: cecilia.summers.07@gmail.com
Please note that I receive a lot of e-mail and may not be able to respond to all messages.