# Javier Turek

> Sr Research Scientist in AI


## Posts

- [NeurIPS 2019 Workshop on Context and Compositionality](https://javierturek.com/neurips-2019-workshop-context-and-compositionality/): We were noticed that our workshop proposal was accepted! We are excited to organize the Context and Compositionality in Biological and Artificial Neural Systems next December at NeurIPS 2019, Vancouver, Canada. During the workshop we will explore advances in context in language processing and their relationship with compositionality from the different angles of Neuroscience, Linguistics/NLP, and Machine Learning. Therefore, we invite researchers in these and subjecent fields to submit their work until September 18th (11:59pm PT). Check the Call for Papers for more information. We are looking forward to your interesting contributions and discussions!
- [NeurIPS 2019 Workshop Submitted](https://javierturek.com/neurips-2019-workshop/): The past year, I have been steering my research to a new fascinating topic: language processing and the brain. Together with Prof. Alex Huth and his PhD student Shailee Jain, we are exploring how the brain processes and utilizes context for language understanding. We come up with the idea to extend our internal discussions to a much broader community that includes researchers from Neuroscience, Linguistics, and Machine Learning. One of the best places where researchers from all these fields congregate is at the NeurIPS conference. Therefore, we embarked in recruiting an amazing group of people that is both interested in […]
- [New Machine Learning Method for Learning about Cognition](https://javierturek.com/machine-learning-method-cognition/): In the past few months, I’ve been collaborating with researchers from the Turk-Browne Lab at Yale University. Their ongoing work is about learning the origins of cognition in the human brain. Equipped with fMRI scanners, they scan kids to analyze their cognitive skills at different ages. Their proposal is simple but quite challenging. The challenges start by recruiting families, making sure they are safe and comfortable during the experiments, developing tasks that are suitable for kids of very young ages, and overcoming the data challenges. In particular, the latter requires to rethink machine learning methods that neuroscientists typically use for analyzing […]
- [A New Method to Approximate Symmetric Matrices](https://javierturek.com/nystrom-symmetric-sparse-approximation/): Symmetric matrices are very common in many fields. For example, they are used in Kernel Machines to  maintain pairwise kernel functions, while in computer vision they represent pairwise distances between points. When a dataset contains ten thousands or more points, these symmetric matrices do not fit in memory and may be too expensive to compute. Existing alternatives suggest to approximate this matrix using a low-rank approximation. Nyström is a very powerful method that samples a subset of the data points and uses them to approximate the matrix. Many research has centered around theory, sampling schemes, and accuracy improvements. However, the […]
- [Learning a Functional Alignment Model with a Semi-Supervised Approach](https://javierturek.com/learning-functional-alignment-semi-supervised/): A few days ago, our paper “A Semi-supervised Method for Multi-Subject fMRI Functional Alignment” was accepted to the IEEE International Conference on Acoustics, Speech and Signal Processing that will be held next March in New Orleans, Louisiana, USA. This work presents an extension to the original Shared Response Model (SRM), an unsupervised method for multi-subject functional alignment of fMRI data. Using a semi-supervised approach, we show how to train SRM taking into consideration data from a supervised task (multi-label classification). In this way, we need almost half the number of unlabeled samples to achieve the same accuracy level, or achieve higher accuracy with the […]
- [Pushing the Limits of Neuroscience](https://javierturek.com/pushing-limits-neuroscience-with-machine-learning/): Neuroscientist is the science of learning how the brain works and understanding, among other things, how the brain stores and processes all the information that is received from the world around it. Several imaging techniques have been developed in recent years that allow neuroscientists to peek inside the human brain. The most important step on this direction is the functional Magnetic Resonance Imaging, or fMRI, that captures the brain activation indirectly from the blood oxygenation levels. With fMRI we can capture a full brain scan every few seconds. Such scans are volumes of the brain comprised of thousands-to-millions of voxels. Processing […]
- [New Paper with Acceleration Framework for Sparse Optimization Problems](https://javierturek.com/sparse-optimization-inverse-covariance-logistic-regression/): I added the link to the paper “A Multilevel Framework for Sparse Optimization With Application to Inverse Covariance Estimation and Logistic Regression” soon to appear in SIAM Scientific Computing (SISC) journal. The paper describes a method that accelerates sparse optimization methods that use L1 regularization to achieve sparse solution. We show how to apply this method to the sparse inverse covariance method (also known as GLASSO) and the L1-regularized logistic regression.
- [Source Code for MAP and MMSE Co-sparse Analysis Model](https://javierturek.com/source-code-map-mmse-co-sparse-analysis-model/): Thanks to some requests that I received lately, I decided to upload source code for the work “On MAP and MMSE Estimators for the Co-sparse Analysis Model”. Hope that this could be useful for more researchers. The code is in github publicly available. Please, I would like to hear you back from everyone using it.
- [New work presented at Optimization workshop at NIPS 2015](https://javierturek.com/new-work-presented-at-optimization-workshop-at-nips-2015/): At the beginning of November our “A Multilevel Acceleration for l1-regularized Logistic Regression” work on how to accelerate the L1-regularized logistic regression problem was accepted to the Optimization workshop at NIPS 2015. Last week, I presented the work in the Optimization workshop at NIPS 2015. This year the Optimization workshop grew a lot, having about 50 posters in several optimization topics. This work was a collaboration between Earn Treister (Univ. Of British Columbia) and myself (Intel Labs).
- [New Journal Paper on Clutter Mitigation using MCA -- Update](https://javierturek.com/clutter-mitigation-mca-ultrasound/): A recent paper “Clutter Mitigation in Echocardiography using Sparse Signal Separation” has been accepted for publication. The article discuss how to apply a sparsity prior to separate clutter from tissue in cardiac ultrasound images. The suggested method uses an adaptive dictionary learned from the patient data using K-SVD. The main challenge of this work was to separate the tissue and the clutter atoms as the trained dictionary includes atoms from both signals. A good separation of the dictionary yields a state-of-the-art clutter mitigation. We tested the robustness of the method and demonstrated its capabilities in real-world sequences. In incoming weeks, the article will be published […]
- [Got the PhD!](https://javierturek.com/got-the-phd/): Last month, I finished my PhD studies and from a few days ago I am a Doctor in Philosophy. My dissertation can be found in the Theses webpage of the Department of Computer Science website from the Technion. The dissertation describes several ways to exploit sparsity as prior information for signal modeling, for signal processing applications, and for parameter estimation.
- [OMP and K-SVD for Complex valued signals](https://javierturek.com/omp-ksvd-complex-signals-mca/): In the works “Clutter Mitigation in Echocardiography using Sparse Signal Separation”, “Sparse Signal Separation with an Off-line Learned Dictionary for Clutter Reduction in Echocardiography“, and “Fusion of Ultrasound Harmonic Imaging with Clutter Removal Using Sparse Signal Separation“, we implemented the algorithm using an Orthogonal Matching Pursuit (OMP) and K-SVD version that works with complex valued signals. I have released the code in the software section. This code is based on the toolboxes published by Dr. Ron Rubinstein and work with Matlab. We used this code to compute sparse representations for signals with phase that were acquired from an ultrasound scanner. You are welcome to […]
- [ICASSP 2015 Paper Accepted](https://javierturek.com/icassp-2015-ultrasound-mca/): Today I received the announcement that the paper “Fusion of Ultrasound Harmonic Imaging with Clutter Removal Using Sparse Signal Separation” was accepted for a presentation in the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2015). The work introduces a novel idea on how speckle noise can be reduced by using a fusion of the fundamental and 2nd harmonics acquired simultaneously. The idea is to remove clutter artifacts while fusing the two harmonic signals. We base the solution on our previous work on clutter mitigation using MCA and the idea of joint sparsity. The method results in improved images both in clutter mitigation and speckle […]
- [Released Code of the Block-Coordinate Descent for Inverse Covariance](https://javierturek.com/code-block-coordinate-descent-inverse-covariance/): I released the code for the paper “A Block-Coordinate Descent Approach for Large-scale Sparse Inverse Covariance Estimation” that was presented in NIPS 2014. The algorithm includes a flag that enables the multilevel acceleration. This flag is very useful for large-scale problems on the thousands-millions of variables. The code runs in Matlab and includes some functions in C that require compilation. Also, it calls functions from METIS 5.0.2 to partitioning the neighbors in every sweep. The released version was tested on Windows, although it should work on other platforms as well. You are welcome to try it and contact me with any comment you may have. I would […]
- [Sparse Inverse Covariance Estimation Paper Published in NIPS Site](https://javierturek.com/sparse-inverse-covariance-estimation-paper-published-nips-site/): Today, I found that the work  “A Block-Coordinate Descent Approach for Large-Scale Sparse Inverse Covariance Estimation” joint with Eran Treister was published in the NIPS 2014 proceedings website. I will publish the algorithm code for this work and the Multilevel framework in a few days. Hope that you enjoy it and please send me your comments!
- [Optimization Workshop OPT2014 in NIPS](https://javierturek.com/optimization-workshop-nips-2014/): Last week, I received the notice that the work with Eran Treister and Irad Yavneh was accepted in the optimization workshop at NIPS 2014. This is a follow up work the sparse inverse covariance work, where we present an acceleration framework based on multilevel techniques. The framework reduces the number of computations needed by defining an hierarchy of levels and updating a subset of the active set of non-zero elements. We tested the framework on QUIC and on BCD-IC algorithms with very interesting results, in particular for large-scale problems where the timings are reduced up to 10x. See you at NIPS 2014 […]
- [NIPS 2014 - Accepted!](https://javierturek.com/nips-2014-accepted/): A few days ago, I’ve received the notification about the acceptance to NIPS 2014 of the work I submitted with my friend and colleague Eran Treister back in June. The NIPS 2014 conference will be held in Montreal, Canada during December 8th and 11th.  Our work is about a new algorithm to solve the Sparse Inverse Covariance Estimation problem in high dimensions, such that the memory is a limitation factor. In the work we show that the algorithm is faster than the previous methods in thousands to millions of variables, and that the algorithm is capable of running in a single server with 64GB because of […]
- [See you in IEEE Israel 2014](https://javierturek.com/ieee-israel-2014/): Our work on Sparse Signal Separation for Clutter Reduction in Echocardiography using Off-line Learned Dictionaries was accepted to be presented in IEEE 28th Convention of Electrical and Electronics Engineers in Israel. The conference will be held in Eilat during December, 2014. The work is about removing clutter artifacts from ultrasound images using sparse representations, morphological component analysis, and off-line dictionary learning.
- [2nd Prize in the CS Research Day 2014](https://javierturek.com/2nd-prize-in-the-cs-research-day-2014/): The poster that I presented about the new work with my friend and colleague Eran Treister, obtained the 2nd place in the CS Faculty Research Day. The event was held last Monday at CS faculty building. Among visitors there were undergrad students, professors, and industry people. The work presents a state-of-the-art method to compute the sparse inverse of the covariance matrix in huge dimensions (hundred thoudsands elements). The method allows for computation of a 100K by 100K matrix in about 10 hours in a quad core computer with 8Gb memory.

## Pages

- [Contact Information](https://javierturek.com/contact-information/): These are the different ways to contact me: Email: javier.turek [at] intel.com   Or if you prefer, using the following form
- [Module Licenses](https://javierturek.com/license-module-pelepay-prestashop/): License for Pelepay Module for Prestashop Copyright 2014 Javier Turek All rights reserved. IMPORTANT-READ CAREFULLY: This End-User License Agreement (”EULA”) is a legal agreement between you (either an individual or a single entity) and Copyright Holder for the “Pelepay Prestashop Module” that accompanies this EULA, which includes computer software and may include associated media, printed materials, “online” or electronic documentation, and Internet-based services (”Software”). An amendment or addendum to this EULA may accompany the software. YOU AGREE TO BE BOUND BY THE TERMS OF THIS EULA BY INSTALLING, COPYING, OR OTHERWISE USING THE SOFTWARE. IF YOU DO NOT AGREE, DO […]
- [Software](https://javierturek.com/software/): Brain Imaging Analysis Kit (BrainIAK) [Source code: GitHub] Block-Coordinate-Descent for Sparse Inverse Covariance Estimation (including Multilevel acceleration) [Download][Old Version] MAP and MMSE methods for the Co-sparse Analysis Model [Source code: GitHub] COMP Toolbox v10 – Implementations of the Orthogonal Matching Pursuit (OMP) and K-SVD algorithms for use with complex signals. [Download] This package extends Dr. Rubinstein’s code in the OMP and the K-SVD Toolboxes [Source].
- [Photography](https://javierturek.com/photography/): In recent years, I started enjoying photography as a personal hobby. Some of the pictures that I take are randomly changing in this website. Enjoy them!  
- [Links](https://javierturek.com/links/): Academic and Collaborators Intel Labs Huth Lab at UT Austin Princeton Neuroscience Institute Turk-Browne Lab at Yale Michael (Miki) Elad’s Home Page Irad Yavneh’s Home Page Eran Triester’s Home Page Jere Sulam’s Home Page Department of Computer Science at the Technion Technion – Israel Institute of Technology
- [Publications](https://javierturek.com/publications/): Publications R. Akiva-Hochman, S. E. Finder, J. S. Turek, E. Treister, “Searching for N:M Fine-grained Sparsity of Weights and Activations in Neural Networks”, European Conference on Computer Vision (ECCV), 2022. [Paper] R. Antonello, V. Vo, J. S. Turek, A. Huth, “Low-Dimensional Structure in the Space of Language Representations is Reflected in Brain Responses”, 35th Conference on Neural Information Processing Systems (NeurIPS 2021), Dec. 2021. [Paper] M. Kumar, M. Anderson, J. Antony, C. Baldassano, P. Brooks, M. Cai, P.-H. Chen, C. Ellis, G. Henselman-Petrusek, D. Huberdeau, B. Hutchinson, Y. Li, Q. Lu, J. Manning, A. Mennen, S. Nastase, H. Richard, A. Schapiro, N. Schuck, M. Shvartsman, N. Sundaram, D. Suo, J. Turek, D. Turner, V. Vo, G. Wallace, Y. Wang, J. Williams, H. Zhang, X. Zhu, M. Capota, J. Cohen, U. Hasson, K. Li, P. Ramadge, N. Turk-Browne, T. Willke, K. Norman, “BrainIAK: The Brain […]
- [Projects](https://javierturek.com/projects/): My Projects Super-resolution with Optical Flow [with Eran Treister, Report] Automatic Cropping of Tagged Areas in Scanned Images [Report] Robot Pursuit [as part of the course Introduction to Robotics 236927, Videos 1 2 3 4 5, Presentation] Projects that I mentored Photo Duel on Android [Students: Marom Sabag, Evgeny Moroshko] Photo Duel on iPhone [Student: Nir Aga] Sparse Coding and Dictionary Learning on GPU [Students: Dor Cohen, Miki Miraz] Quaternion K-SVD for Color Image Denoising [Student: Amit Carmeli]  
- [Teaching](https://javierturek.com/teaching/): During the past 6 years, I taught several programming courses at the Department of Computer Science at the Technion. Lecturer 234112 Introduction to Programming in C (Spring 2013/14) 234127 Introduction to Matlab (Summer 2013, Winter 2013/14) 234114 Introduction to Computer Science M/H (Spring 2012/13) Teaching Assistant in Charge 234114 Introduction to Computer Science M/H (TA in charge: Winter 2011/12, Spring 2011/12) Teaching Assistant 234112 Introduction to Programming in C (Spring 2010/11, TA in charge: Winter 2011/12, Spring 2011/12, Lecturer: Spring 2012/13) 234114 Introduction to Computer Science M/H (Spring 2010/11) 234122 Introduction to Systems Programming (Spring 2007/08)  
- [Personal Information](https://javierturek.com/): I am a Research Scientist at a startup in stealth mode.  My objective is to disrupt one of the oldest industries in the world through innovation with applied AI. Currently, my work focuses on improving neural networks for language modeling and perception. Before my current role, I was a Staff Research Scientist at Intel Labs. I have been part of amazing academic collaborations with Neuroscientists and Computer Scientists. I learned to analyze fMRI data, to understand neuroscience challenges, to discover the gaps with the human mind, and to consider collaboration with humans as the key aspect of future AI systems. […]


[comment]: # (Generated by Hostinger Tools Plugin)