Sina Honari

Home | Research | Awards | Hobbies

---

I am a senior research scientist at the Samsung AI Center Toronto. Before joining Samsung, I was a postdoctoral researcher in the computer vision lab (CVLab) at EPFL in Lausanne-Switzerland, working with Pascal Fua and Mathieu Salzmann for more than 3 years. I completed my PhD in 2019 at Mila and the University of Montreal under the supervision of Christopher Pal and Pascal Vincent. After my PhD, I held a 6-month postdoctoral position at Mila and Polytéchnique Montréal with Christopher Pal.

My broad research domains are developing machine learning and deep learning techniques for computer vision and vision-language problems. In particular, I am interested in developing semi-supervised, self-supervised, unsupervised, and out-of-domain generalization learning techniques, by leveraging task-related, computer-vision, or auxiliary priors. More recently, I have been exploring the use of large language and vision-language models for image quality analysis and perception. The application of my projects has been in image restoration and enhancement, low-light image denoising, image super-resolution, image quality analysis, multi-view geometry, 2D landmark localization, 3D human pose estimation, medical image segmentation and domain adaptation, bias-mitigation analysis, image generation, visual question answering, animal action recognition, and road-obstacle and anomaly detection.

During my studies, I interned at Morgan Stanley as a trade technology analyst and at NVIDIA’s Learning & Perception Research team (led by Jan Kautz) as a deep learning researcher. My PhD was partially funded by Fonds de Recherche du Québec - Nature et Technologies (FRQNT).