Eugene Brevdo

Former Ph.D. Candidate (now at Google)
Graduated: July 2011

Email: ebrevdo at gmail



Curriculum Vitae

About Me

I was a PhD candidate in the Department of Electrical Engineering at Princeton University. My advisers were Peter Ramadge and Ingrid Daubechies. My interests include neural networks, statistical signal processing, geometry of learning and manifold learning, harmonic analysis, compressive sensing, and their applications to inverse problems, biomedical imaging, and forgery detection in art.


My PhD thesis is available for download:

Efficient Representations of Signals in Nonlinear Signal Processing with Applications to Inverse Problems (pdf)


Instantaneous Frequency Estimation and Applications

Click here to go to the homepage of the MATLAB Synchrosqueezing Toolbox.

The Synchrosqueezing algorithm for time-varying spectral analysis: robustness properties and new paleoclimate applications.
G. Thakur, E. Brevdo, N.S. Fučkar, and H-T. Wu. Submitted, July 2012. (arXiv) (code)

Nonlinear Signal Processing and Machine Learning

Semisupervised Learning: The Regularized Laplacian, Geodesics, and the Small Viscosity Limit.
E. Brevdo and P. J. Ramadge. Submitted, April 2011.

Bridge Detection and Robust Geodesics Estimation via Random Walks.
E. Brevdo and P. J. Ramadge. International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Dallas, TX, March 2010. (paper) (poster)

Image Processing for Artist Identification: Stylistic Analysis

Stylistic Analysis of Paintings Using Wavelets and Machine Learning.
S. Jafarpour, G. Polatkan, E. Brevdo, S. Hughes, A. Brasoveanu, and I. Daubechies. European Signal Processing Conference (EUSIPCO) 2009. (paper)

Image Processing for Artist Identification: Brushwork in the Paintings of Vincent Van Gogh.
C. R. Johnson, Jr., E. Hendriks, I. Berezhnoy, E. Brevdo, S. Hughes, I. Daubechies, J. Li, E. Postma, and J.Z. Wang. IEEE Signal Processing Magazine Special Issue on Visual Cultural Heritage, July 2008. (link)