Setareh Cohan

I am a Ph.D. student in computer science at the University of British Columbia, supervised by Michiel van de Panne. My research goal is to study machine learning for efficient decision making. I am currently interested in learning-based methods for control including reinforcement learning and generative models.

Prior to that, I completed my M.Sc. in computer science at the University of British Columbia, supervised by Jim Little and Leonid Sigal.

I did my B.Sc. in computer engineering at Sharif University of Technology.

Email  /  CV  /  Github

Publications
Flexible Motion In-betweening with Diffusion Models
Setareh Cohan, Guy Tevet, Daniele Reda, Xue Bin Peng, Michiel van de Panne
SIGGRAPH 2024 [Arxiv][Project Page][Code]
Understanding the Evolution of Linear Regions in Deep Reinforcement Learning
Setareh Cohan, Nam Hee Kim, David Rolnick, Michiel van de Panne
NeurIPS 2022 [Arxiv][Project Page][Code]


Projects
Diffusion Models for Motion Synthesis
Setareh Cohan
Ongoing project in which we explore the capabilities of probabilistic diffusion models for human motion synthesis and editing.
A Bayesian Approach to Visual Question Answering
Gursimran Singh, Saeid Naderiparizi, Setareh Cohan
Probabilistic Programming course, winter 2018 [PDF]
RegNet: Regularizing Deep Networks
Setareh Cohan, Saeid Naderiparizi
Multimodal Learning with Vision, Language and Sound course, spring 2018 [PDF] [code]
A Survey on Active Learning
Setareh Cohan, Saeid Naderiparizi
Machine Learning II course, spring 2018 [PDF]
Learning to Ride a Bike
Setareh Cohan, Saeid Naderiparizi
Computer Animation course, winter 2017 [PDF] [code]
Learning Influence Diffusion Probabilities under Independent Cascade with Independent Decay over Time
Saeid Naderiparizi, Polina Zablotskaia, Setareh Cohan
Social and Information Networks course, winter 2017 [PDF]
3D Scene Reconstruction
Setareh Cohan, Saeid Naderiparizi
Fundamentals of 3D Computer Vision course, spring 2015 [PDF]


Experience
Research Intern at Borealis AI
Supervisor: Lili Meng
Investigated the advantages of integrating probabilistic prediction and selective prediction (or prediction with reject option) for regression tasks. Implemented SelectiveNet (Geifman et al., ICML2019), and altered the prediction modules to probabilistic prediction modules (code is available here). Designed and executed numerical experiments on large computer clusters for classification and regression tasks.


Teaching

Teaching Assistant for CPSC 440, Advanced Machine Learning - Winter 2020
Teaching Assistant for CPSC 340, Machine Learning - Winter 2019
Teaching Assistant for CPSC 304, Database Design - Fall 2018
Teaching Assistant for CPSC 320, Algorithm Design and Analysis - Fall 2019, Summer 2019, Winter 2018, Fall 2017
Teaching Assistant for Design of Algorithms - Winter 2016
Teaching Assistant Advanced Programming in Java - Spring 2014



Education
ubc

University of British Columbia 2020 - Present
Ph.D. in Computer Science
Supervisor: Michiel van de Panne

ubc

University of British Columbia 2017 - 2019
M.Sc. in Computer Science
Supervisors: Jim Little, Leonid Sigal

ubc

Sharif University of Technology 2012 - 2017
B.Sc. in Computer Engineering

ubc

Farzanegan High School 2008 - 2012
Diploma in Mathematics and Physics Discipline


Website source from Jon Barron.