Parvaneh Saeedi, Ph.D., P.Eng.
Email: psaeedi@sfu.ca
Phone: +1 (778) 782-4746
School of Engineering Science
51社区黑料
8888 University Dr., Burnaby
BC, V5A 1S6, Canada
Office: ASB10837
Email: psaeedi@sfu.ca
Phone: +1 (778) 782-4746
School of Engineering Science
51社区黑料
8888 University Dr., Burnaby
BC, V5A 1S6, Canada
Office: ASB10837
Dr. Parvaneh Saeedi is a Professor of Engineering Science with expertise in artificial intelligence, computer vision, and medical image analysis. Her research focuses on developing robust and interpretable machine learning methods for healthcare, biomedical imaging, and geospatial applications. She has authored many peer reviewed publications in leading journals and conferences, including various IEEE Transactions, Medical Image Analysis, MICCAI, CVPR, and ICIP. Her work has contributed to advances in embryo and fetal health assessment, privacy preserving learning, and remote sensing image understanding.
She served as Associate Dean of Research and Graduate Studies for the Faculty of Applied Sciences at 51社区黑料 from 2019 to 2024, where she provided faculty wide leadership in research planning and graduate education. In that role, she oversaw graduate program development, coordinated closely with university level offices, and addressed complex academic and administrative issues affecting students and supervisors. During her term, Professor Saeedi contributed directly to the design and implementation of new graduate programs and led strategic initiatives that strengthened research capacity and fostered interdisciplinary collaboration across the faculty. Her tenure coincided with the COVID 19 pandemic, during which she served the faculty and school with exceptional dedication and collaborative leadership, navigating rapidly evolving governance and safety requirements while mitigating disruptions to research productivity and graduate training.
Dr. Saeedi has received competitive fundings and awards from national and international agencies and actively collaborates with clinicians, hospitals, and industry partners. She was the recepient of NSERC's University Faculty Award from 2007 to 2012. She is committed to mentoring highly qualified personnel and translating research outcomes into real world impact.
Professor Saeedi is a long-standing advocate for equity and inclusion, having served as Women in Engineering faculty mentor since 2012 and as Chair of Equity & Diversity Committee in the School of Engineering Science, where she actively supports women and underrepresented groups in engineering. She is an avid supporter for animal protection and welfare. Since 2000, she has been actively involved as a Vision Mate with the Canadian National Institute for the Blind (CNIB), contributing to community support and accessibility for individuals living with vision loss in Canada.
Her research focuses on the development of scalable, privacy aware, and robust artificial intelligence systems that operate effectively in real world, data distributed environments. Spanning collaborative intelligence, federated and split learning, medical image and video analysis, and unsupervised representation learning, the work addresses fundamental challenges in deploying AI beyond centralized settings, including communication constraints, system heterogeneity, limited supervision, and privacy risks. By combining principled algorithm design with system level considerations, the research develops methods that enable reliable collaboration across institutions while preserving data confidentiality and model performance. Collectively, these contributions support the deployment of trustworthy and efficient AI technologies in healthcare and other high impact domains, aligned with national priorities in responsible artificial intelligence, health innovation, and equitable access to advanced computational tools.
Collaborative intelligence enables the distribution of artificial intelligence computation across multiple devices and institutions, offering a scalable and privacy aware pathway for deploying medical AI beyond centralized data centers and closer to the point of care. This research focuses on the design of communication efficient, privacy preserving, and robust collaborative intelligence frameworks for medical image analysis. The work addresses key limitations of centralized and conventional federated learning, including data sharing constraints, unreliable communication, client heterogeneity, and deployment in real clinical environments. The main objective is to integrate split learning, federated optimization, feature compression, and error resilience mechanisms to enable scalable and trustworthy AI across distributed healthcare institutions while maintaining patient data privacy.
Video summarization aims to automatically distill long videos into concise representations that preserve the most informative and relevant content, enabling efficient understanding, retrieval, and analysis of large scale video data. In this research work, we are aiming to generate concise and informative summaries of medical data without relying on human annotations. The approach leverages reinforcement learning and transformer based models, using video reconstruction quality as a learned reward signal to identify the most salient frames. By learning directly from raw data, the framework avoids hand crafted heuristics and unstable adversarial training while preserving essential video content.
This research advances principled approaches to biomedical image segmentation by focusing on how models learn reliable and clinically meaningful representations from complex, imperfect, and heterogeneous medical imaging data. By treating segmentation as an adaptive learning process rather than a purely label driven task, the work emphasizes robustness, generalization, and the structured use of feedback to improve boundary awareness and reduce sensitivity to annotation noise and domain variability. The overarching goal is to develop segmentation frameworks that remain reliable across diverse imaging settings while supporting scalable, privacy preserving, and clinically relevant deployment.
Courses taught at 51社区黑料.
I am currently recruiting one PhD student. Applicants must hold a Master鈥檚 degree in Electrical and Computer Engineering with a strong focus on Artificial Intelligence, have a minimum GPA of 80 percent, and demonstrate a strong research record, including peer reviewed publications.
Due to the high volume of inquiries I receive, I am only able to respond to candidates whose background and qualifications closely align with these requirements.