I work with Tom Goldstein in non-convex optimization problems. Currently, I'm trying to solve low rank matrix recovery in the natural parameter space without lifting to higher dimensions. Previously, I worked on Phase Retrieval.

I also answer questions on mathematical optimization on Quora where I am the top writer in this category.

NEW!  We created PhasePack- a MATLAB based library to solve phase retrieval using classical and contemporary routines. Get the Pack here!

NEW!  Our project on texture synthesis using deep learning is complete! Check out the result below!   Report


  • Phasepack User Guide
    Rohan Chandra, Ziyuan Zhong, Justin Hontz, Val McCulloch, Christoph Studer, Tom Goldstein, 2017
    arXiv preprint
  • Phasepack: A Phase Retrieval Library
    Rohan Chandra, Ziyuan Zhong, Justin Hontz, Val McCulloch, Christoph Studer, Tom Goldstein
    Submitted to IEEE Proceedings of the 51st Asilomar Conference on Signals, Systems and Computers.
  • Multiply by 9
    Arthur Benjamin, Rohan Chandra
    The College Mathematics Journal.

Current Projects

Low Rank Matrix Estimation Without Lifting - As part of my MS thesis, I am trying to solve the problem of low rank matrix recovery without lifting i.e. in the natural parameter space. Initial sketches of the proof look promising and I am currently in the process of formalizing this.
Texture Synthesis - Generating textures using deep learning. Results,  Report
PhasePack - A MATLAB based library for solving Phase Retrieval.

Past Projects

Autonomous Cars - Developed a real-time lane detection module. Video
Structure from Motion - Implemented the full pipeline to reconstruct the 3D scene from multiple images.  Code
Semantic Mapping of 3D pointcloud - Given images of single and muliple cluttered table-top objects, generate 3-D pointclouds and do the following: Video 1 Video 2 Code
  • Construct 3-D model of the scene
  • 3D pointcloud Segmentation
  • Generating a semantic map

Relevant Courses

MATH 461: Undergraduate Linear Algebra
CMSC 828G: Graduate Computer Vision I by Rama Chellapa
CMSC 733: Graduate Computer Vision II by Yiannis Aloimonos
CMSC 764: Graduate Optimization by Tom Goldstein
STAT 400: Probability and Statistics
CMSC 726: Graduate Machine Learning by Jordan Boyd-Graber (current)
CMSC 828R: Spectral Methods and Reinforcement Learning by Furong Huang (current)
STAT 420: Statistics I
CMSC 818G: Information-Centric Design of Systems by Ashok Agrawala


Fall 2016: CMSC 417  - Computer Networks (as TA)
Spring 2017: CMSC 131  - Object Oriented Programming (Java) (as TA)
Fall 2017: CMSC 250 - Discrete Structures (as TA)

Department Service

Reviewer for MS applications to the CS department at UMD for Fall 2017 and Fall 2018.