Soumya Ghosh
sghosh@cs.brown.edu

I am a research scientist at IBM research. I work on developing statistical models for explaining images, text and noisy, real world healthcare data. At Brown, I was advised by Erik Sudderth. Before Brown, I spent a few years in beautiful Boulder getting my master's degree from the University of Colorado. At Colorado, I was advised by Jane Mulligan. Going further back, I went to the University of Mumbai (Bombay) (KJSCE) as an undergrad. I also spent a year as a postdoctoral scientist at the now defunct Disney research, Cambridge.
Selected Recent Research



Ajjen Joshi, Soumya Ghosh, Margrit Betke, Stan Scarloff, Hanspeter Pfister
CVPR 2017.
Ajjen Joshi, Soumya Ghosh, Margrit Betke, Hanspeter Pfister
NIPS 2016. Workshop on Bayesian Deep Learning.
  

                
Soumya Ghosh, Zhaonan Sun, Ying Li, Yu Cheng, Amrita Mohan,
Cristina Sampaio, Jianying Hu.

AMIA Joint Summit on Translational Science 2017.

 
                  
Soumya Ghosh, Francesco DelleFave, Jonathan Yedidia
AAAI 2016.

 
                
Soumya Ghosh, Erik Sudderth
NIPS 2015. Workshop on Bayesian Nonparametrics: The Next Generation.

 
               
Soumya Ghosh, Michalis Raptis, Leonid Sigal, Erik Sudderth
UAI 2014.


                 Soumya Ghosh, Erik Sudderth, Matthew Loper, Michael Black
NIPS 2012.


                Soumya Ghosh, Erik Sudderth
CVPR 2012.


                Soumya Ghosh, Andrei Ungureanu, Erik Sudderth, David Blei
NIPS 2011.

Dissertation

              Soumya Ghosh
Brown University, 2015
Recent Talks

                                                             
Learning and inference in distance dependent models.  [slides]
    • Boston University, IVC seminar,  Nov 2015.
Bayesian nonparametric discovery of layers and parts from scenes and objects. [slides]
    • Philips Research, Briarcliff Manor, NY, Jun 2015.
    • IBM Research, Yorktown Heights,  NY, Apr 2014.
    • HP, Palo Alto, CA, Apr 2014.
    • Bosch Research, Palo Alto, CA, May 2014.
    • Exxon Mobil, Corporate Strategic Research, Clinton, NJ, May 2014.
    • BBN Technologies, Cambridge, MA, May 2014.
    • Disney Research, Boston, MA, May 2014
    • Schlumberger-Doll Research center, Cambridge, MA, May 2014.
Statistical models for spatially correlated data. [slides]
    • Microsoft Research, New England,  Jun 2013.
Nonparametric learning for layered segmentation of natural images. [slides]
    • Disney Research, Pittsburgh,  Jun 2012.
Older Work




Unsupervised, semi-supervised
learning, image segmentation and Martian landform analysis





Soumya Ghosh, Jane Mulligan
2010 IEEE International Conference on Robotics and Automation (ICRA 2010)

Soumya Ghosh, Joseph J. Pfeiffer III, Jane Mulligan
2009 IEEE Workshop on Applications of Computer Vision (WACV 2009)

Steven Bethard, Soumya Ghosh, James Martin, Tamara Sumner
2009 Joint Conference on Digital Libraries (JCDL 2009)

Soumya Ghosh, Soundar Srinivasan, Burton Andrews
2009 International Joint Conference on Neural Networks (IJCNN 2009)

Soumya Ghosh, Tomasz Stepinski , Ricardo Vilalta
IEEE Transactions on Geoscience and Remote Sensing. Volume 48,Issue 1, Jan 2010 Tomasz Stepinski, Soumya Ghosh, Ricardo Vilalta
19th Conference on Innovative Applications of Artificial Intelligence (AAAI/IAAI 2007)