Sobhan Naderi Parizi

I obtained my PhD from the School of Engineering at Brown University where I was advised by Prof. Pedro Felzenszwlab

am interested in developing new learning frameworks for various vision tasks with a special focus on reducing dependence on supervision and hand-engineered interventions. In a broader sense, I am passionately interested in understanding what it means to "learn" from "data", what can we learn without supervision, and what signal should unsupervised learning be driven by.


Research Interests

  • Weakly supervised learning
  • (Non-)Convex optimization
  • Latent Variable Models
  • Structured output prediction
  • Learning and structured prediction with deep neural networks


PhD in Computer Vision
Brown University, Providence
Sep 2011 - Jan 2016
MSc in Systems, Control, & Robotics
Royal Institute of Technology (KTH), Stockholm
Sep 2008 - Sep 2010
Modeling and Optimization of Classifiers with Latent Variables
Advised by Pedro Felzenszwalb, pdf, slides.
Reducing Ambiguity of Local Descriptors for Visual Recognition
Advised by Ivan Laptev & Alireza Tavakoli, pdf.

MSc in Computer Science
University of Chicago, Chicago
Sep 2010 - Sep 2011
BSc in Computer Engineering
Tehran Polytechnic (Amirkabir), Tehran
Sep 2003 - Jan 2008
Image Classification with Reconfigurable Spatial Structures
Advised by Pedro Felzenszwalb, pdf.
Adaptive continuous Hidden Markov Models integrated with language grammars
Advised by Mohammad Mehdi Homayounpour, pdf (in persian), slides (in persian).


Generalized Majorization-Minimization
S. Naderi Parizi, K. He, S. Sclaroff, P. Felzenszwalb
arXiv tech report, 2015
Spotlight the Negatives: A Generalized Discriminative Latent Model
H. Azizpour, M. Arefiyan, S. Naderi Parizi, S. Carlsson
BMVC 2015
pdf, poster

Automatic Discovery and Optimization of Parts for Image Classification
S. Naderi Parizi, A. Vedaldi, A. Zisserman, P. Felzenszwalb
ICLR 2015
pdf, poster

Reconfigurable Models for Scene Recognition
S. Naderi Parizi, J. Oberlin, P. Felzenszwalb
CVPR 2012

Improving Bag-of-Features Action Recognition with Non-local Cues
M.M. Ullah, S. Naderi Parizi, I. Laptev
BMVC 2010

Modeling Image Context using Object Centered Grids
S. Naderi Parizi, I. Laptev, A. Tavakoli
DICTA 2009
pdfslides, project

Reading Street Signs Using a Generic Structured Object Detection and Signature Recognition Approach
S. Naderi Parizi, A. Tavakoli, O. Aghazadeh, J.O. Eklundh
pdfslides, project 

N4: A Precise and Highly Sensitive Promoter Predictor Using Neural Network Fed by Nearest Neighbors
A. Askary, A. Masoudi-Nejad, A. Mizbani, R. Sharafi, S. Naderi Parizi, M. Purmasjedi
Journal of Genes and Genetic Systems, 2009

A New Error Correction Code
A.S. Shahmiri, S. Naderi Parizi, M.K. Akbari
ICLAN 2007