Sageev Oore Homepage

Research Interests: Machine Learning & AI

Current projects include:

• Interactive AI, Music & Creativity support: Human-collaborative musical AI systems
(e.g. Summer 2025 blog post, UIST 2023, IUI 2024, ICML 2024 (oral))

• AI for Health: Tracking mental health
(e.g. Nature Sci Reports 2023)

• AI Robustness: Diffusion-based augmentation for robust classification
(e.g. NeurIPS 2024, ICML 2020)

• ML Theoretical Foundations: Riemannian geometry for better predicting generalization abilities of transformer models
(e.g. ICML 2025 (spotlight) )

Many of our projects (e.g. in human health and creativity) are inherently collaborative and interdisciplinary.

Selected previous projects include AI and machine learning for: interactive support for whale tracking, respiratory-assisting technology, robot localization, learning movement representation for patients with Parkinson's Disease, biomedical image segmentation, and high-dimensional interactive real-time control interfaces.

For Prospective Students

I have well-funded openings for motivated graduate and honours undergraduate students with strong backgrounds.
Please also see the important note below about diversity and inclusiveness
If you wish to email me about potential graduate or undergraduate studies, please include the subject line
"for consideration: studies at Dalhousie/Vector".
My contact information begins with "osageev", ends with "gmail.com", and has an 'at' sign in the middle.
Dalhousie University is committed to fostering a collegial culture grounded in diversity and inclusiveness. The University encourages applications from Indigenous people, persons with a disability, racially visible persons, women, and persons of minority sexual orientations and gender identities, and all candidates who would contribute to the diversity of our community.

(old) News

This page is currently out of date; I will be updating it to include... New News!

2020 Outstanding Demonstration Award (Runner-up), NeurIPS, interactive tool for speech-to-music
PMLR paper by D'Eon (*), Dumpala (*), Sastry (*), Oore, and Oore
2020 Best Paper Award, CVPR ISIC Workshop on Medical Skin Image Classification (uses the Gram-OOD method below)
2020 ICML paper: Gram-OOD is a new method for OOD detection giving SOTA results on far-OOD examples
2018 Awarded Canada CIFAR AI Chair
2017 Best Demonstration Award, AAAI, Honorable Mention for interactive system for deep-learning-based musical dialogue
2016 Best Demonstration Award, NeurIPS, interactive music generation system
2016 - 2018 Visiting Research Scientist, Magenta Team at Google Brain Research
PerformanceRNN: a recurrent neural network music generation system,
Autoencoders for Musical Dialogue
Radical Cycle is a CD that I released with my brother, Dani Oore. (Live performance excerpt)
Here is an old video of me using the Glove Talk II system.

Brief Biography

Sageev Oore is a faculty member in Computer Science at Dalhousie University (Halifax), a Faculty Affiliate at the Vector Institute for Artificial Intelligence (Toronto), and a former Canada CIFAR AI Chair. He is interested in basic research in machine learning and deep learning, and building effective AI systems for supporting health and human creativity. During 2016--2018 he worked as Visiting Scientist in Google Brain Research (California) on the Magenta team, developing generative music systems such as PerformanceRNN. Sageev is an award-winning musician; as a pianist he has performed as soloist with orchestras both as a classical soloist and as an improviser. Sageev studied piano performance with music teachers from Dalhousie, Juilliard, and UBC. He completed his undergraduate degree in Mathematics (Dalhousie), and MSc and PhD degrees in Computer Science (University of Toronto) working with Geoffrey Hinton.