Rohan Varma

Rohan Varma

Auto-generated profile, based on publicly available information

Rohan Varma

Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, USA

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 March 2016
Verified
Conference Paper
Closed access
SCSiheng Chen
Siheng Chen
RVRohan Varma
Rohan Varma
ASAarti Singh
Aarti Singh
JKJelena Kovacevic
Jelena Kovacevic
We study representations of piecewise-smooth signals on graphs. We first define classes for smooth, piecewise-constant, and piecewise-smooth graph signals, followed by a series of multiresolution local sets to analyze those signals by implementing a multiresolution analysis on graphs. Based on these local sets, we propose local-set-based piecewise-constant and piecewise-smooth dictionaries as graph signal representations that, in spirit, resemble the classical Haar wavelet basis and are naturally localized in both graph vertex and graph Fourier domains. Moreover, they promote sparsity when representing piecewise-smooth graph signals. In the experiments, we show that local-set-based dictionaries outperform graph Fourier domain based representations when approximating both simulated and real-world graph signals.

1 December 2017
Verified
Conference Paper
Closed access
RVRohan Varma
Rohan Varma
SCSiheng Chen
Siheng Chen
JKJelena Kovacevic
Jelena Kovacevic
In this paper, we study the recovery of the graph topology or structure. We first extend our previous work on graph signal recovery to present a joint graph signal and structure recovery framework. By doing this, we allow the algorithm to learn a graph structure from noisy and incomplete graph signals and recover the graph signals at the same time. In this paper, we particularly focus on the specific subproblem of graph structure learning and develop algorithms towards this problem and analyze them. We briefly study the implications when the underlying true graph structure is irregular or regular. Finally, we validate the proposed methods for both synthetic data and the real-world recovery problem of semi-supervised digit-image classification.