Paris Smaragdis

Senior Research Scientist

Creative Intelligence Lab, San Francisco

Paris is a part of the faculty at the University of Illinois and is a part-time research scientist at Adobe. His broader area of interest encompasses anything that involves sound. He spent most of his time working on novel uses of machine learning and signal processing on audio signals, and thinking on what it takes to make a machine listen, understand and appreciate the expressiveness of sound the same way we do.

He completed his Ph.D. and S.M. degrees and did a postdoc at the MIT Media Lab while working at the Machine Listening Group under professor Barry Vercoe. He was also very fortunate to spend time working at some amazing research labs such as Interval Research, Starlab, and Mitsubishi Electric Research Laboratories (MERL). In 2006 he was honored by the MIT Technology Review which recognized him as one of the year’s top 35 young technology innovators.

My Publications

Blind Multi-Channel Source Separation by Circular-Linear Statistical Modeling of Phase DifferencesPhysics Storyboards

Traa, J., Smaragdis, P. (May. 1, 2013)
ICASSP 2013 – Proceedings of the 38th International Conference on Acoustics, Speech, and Signal Processing, May 2013

Collaborative Audio Enhancement Using Probabilistic Latent Component Sharing

Kim, M., Smaragdis, P. (May. 1, 2013)
ICASSP 2013 – Proceedings of the 38th International Conference on Acoustics, Speech, and Signal Processing, May 2013

Prediction Based Filtering and Smoothing to Exploit Temporal Dependencies in NMF

Mohammadiha, N., Smaragdis, P., Leijon, A. (May. 1, 2013)
ICASSP 2013 – Proceedings of the 38th International Conference on Acoustics, Speech, and Signal Processing, May 2013

Clustering and Synchronizing Multi-camera Video via Landmark Cross-correlation

Bryan, N., Smaragdis, P., Mysore, G. (Mar. 25, 2012)
ICASSP - IEEE International Conference on Acoustics, Speech, and Signal Processing , March 2012

Following Musical Sources by Example

Smaragdis, P., Mysore, G. (Mar. 25, 2012)
ICASSP - IEEE International Conference on Acoustics, Speech, and Signal Processing , March 2012

Noise-Robust Dynamic Time Warping Using PLCA Features

King, B., Smaragdis, P., Mysore, G. (Mar. 25, 2012)
ICASSP - IEEE International Conference on Acoustics, Speech, and Signal Processing , March 2012

A Non-negative Approach to Language Informed Speech Separation

Mysore, G., Smaragdis, P. (Mar. 12, 2012)
LVA/ICA - International Conference on Latent Variable Analysis and Signal Separation, March 2012

Online PLCA for Real-Time Semi-supervised Source Separation

Duan, Z., Mysore, G., Smaragdis, P. (Mar. 12, 2012)
LVA/ICA - International Conference on Latent Variable Analysis and Signal Separation, March 2012

Sound Recognition in Mixtures

Nam, J., Mysore, G., Smaragdis, P. (Mar. 12, 2012)
LVA/ICA - International Conference on Latent Variable Analysis and Signal Separation, March 2012

A Convolutive Spectral Decomposition Approach to the Separation of Feedback from Target Speech

Mysore, G., Smaragdis, P. (Sep. 18, 2011)
MLSP - IEEE International Workshop on Machine Learning for Signal Processing , September 2011

A Non-Negative Approach to Semi-Supervised Separation of Speech from Noise with the Use of Temporal Dynamics

Mysore, G., Smaragdis, P. (May. 22, 2011)
ICASSP - IEEE International Conference on Acoustics, Speech, and Signal Processing , May 2011

An Adaptive Time-Frequency Resolution Approach for Non-Negative Matrix Factorization Based Single Channel Sound Source Separation

Kirbiz, S., Smaragdis, P. (Feb. 1, 2011)
ICASSP 2011 - International Conference on Acoustics, Speech, and Signal Processing

Nearest Subspace Representations for Sound Mixtures

Smaragdis, P. (Feb. 1, 2011)
ICASSP 2011 - International Conference on Acoustics, Speech, and Signal Processing

Non-negative Hidden Markov Modeling of Audio with Application to Source Separation

Mysore, G., Smaragdis, P., Raj, B. (Sep. 27, 2010)
Best Student Paper Award LVA/ICA - International Conference on Latent Variable Analysis and Signal Separation, September 2010

Missing Data Imputation for Time-Frequency Representations of Audio Signals

Smaragdis, P., Raj, B., Shashanka, M. (Aug. 30, 2010)
In the Journal of Signal Processing Systems . August 2010

The Markov Selection Model For Concurrent Speech Recognition

Smaragdis, P., Raj, B. (Feb. 2, 2010)
In IEEE International Workshop on Machine Learning for Signal Processing, 2010

Automatic Identification of Individual Killer Whales

Brown, J., Smaragdis, P., Nousek-McGregor, A. (Feb. 1, 2010)
In The Journal of the Acoustical Society of America, 128

A Sparse Non-parametric Approach for Single Channel Separation of Known Sounds

Smaragdis, P., Shashanka, M., Raj, B. (Dec. 30, 2009)
In Neural Information Processing Systems. Vancouver, BC, Canada . December 2009

User Guided Audio Selection from Complex Sound Mixtures

Smaragdis, P. (Oct. 30, 2009)
In the 22nd ACM Symposium on User Interface Software and Technology (UIST 09) . Victoria, BC, Canada. October 2009

Separation by Humming: User Guided Sound Extraction from Monophonic Mixtures

Mysore, G., Smaragdis, P. (Oct. 18, 2009)
WASPAA - IEEE Workshop on Applications of Signal Processing to Audio and Acoustics , October 2009

Dynamic Range Extension Using Interleaved Gains

Smaragdis, P. (Jul. 30, 2009)
In IEEE Transactions of Audio, Speech and Language Processing. July 2009

Multipitch Estimation using Sparse Impulse Distributions and Instrument Specific Priors

Mysore, G., Smaragdis, P. (Jun. 18, 2009)
IMCL - International Conference on Machine Learning (ICML) Workshop on Sparse Methods for Music Audio , June 2009

Hidden Markov and Gaussian Mixture Models for Automatic Call Classification

Brown, J., Smaragdis, P. (May. 30, 2009)
In Journal of the Acoustical Society of America Express Letters. May 2009

Relative Pitch Tracking of Multiple Arbitrary Sounds

Smaragdis, P. (May. 30, 2009)
In Journal of the Acoustical Society of America , Volume 125, Issue 5, pp. 3406-3413. May 2009

Probabilistic Factorization of Non-Negative Data with Entropic Co-occurrence Constraints

Smaragdis, P., Shashanka, M., Raj, B., Mysore, G. (Mar. 15, 2009)
ICA - International Conference on Independent Component Analysis and Signal Separation , March 2009

Sparse and Shift-invariant Feature Extraction from Non- negative Data

Smaragdis, P., Raj, B., Shashanka, M. (Dec. 18, 2008)
In Proc. of the IEEE International Conference on Audio and Speech Signal Processing (ICASSP)

Blind Speech Separation

Smaragdis, P. (Dec. 10, 2007)
eds. S. Makino, T. L. and Sawada, H.. Springer, ch. Component Based Techniques for Monophonic Speech Separation and Recognition

Sparse Overcomplete Decomposition for Single Channel speaker Separation

Shashanka, M., Raj, B., Smaragdis, P. (Nov. 17, 2007)
In Proc. of the IEEE International Conference on Audio and Speech Signal Processing (ICASSP)

Example-driven Bandwidth Expansion

Smaragdis, P., Raj., B. (Oct. 30, 2007)
In Proceedings of IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) . New Paltz, NY. October 2007

Sparse Overcomplete Latent Variable Decoposition of Counts Data

Shashanka, M., Raj, B., Smaragdis, P. (Jun. 25, 2007)
In Proc. of Neural Information Processing Systems (NIPS)

A Framework for Secure Speech Recognition

Smaragdis, P., Shashanka, M. (May. 28, 2007)
IEEE Transaction on Audio, Speech and Language Processing , 4, 1404–1413

Position and Trajectory Learning for Microphone Arrays

Smaragdis, P., Boufounos, P. (May. 17, 2007)
IEEE Transaction on Audio, Speech and Language Processing , 15, 358–368

Convolutive Speech Bases and Their Application to Speech Separation

Smaragdis, P. (Mar. 20, 2007)
IEEE Transaction on Audio, Speech and Language Processing , 15, 1–12

Component Based Techniques for Monophonic Speech Separation and Recognition. In “Blind Speech Separation”

Smaragdis, P. (Jan. 30, 2007)
S. Makino, T-W.Lee and H. Sawada (eds.) Blind Speech Separation, Springer

A probabilistic latent variable model for acoustic modeling

Smaragdis, P., Raj, B., Shashanka, M. (May. 29, 2006)
In Neural Information Processing Systems (NIPS) Workshops