Victor Fragoso victor.fragoso [at] microsoft.com

I am a Researcher at Microsoft working on Computer Vision problems. I received my PhD from the University of California Santa Barbara in Dec. 2014 under the supervision of Prof. Matthew Turk. Before joining Microsoft, I served as an Assistant Prof. at WVU.

Research Overview

I am interested in statistical modeling and scalable computational techniques for computer vision and machine learning. My aim is to create theoretically sound, robust, and scalable computer vision and machine learning algorithms. I am interested in designing self assessment mechanisms that can reason about the correctness of a decision outcome given a noisy input, e.g., predicting the correctness of a nearest-neighbor classification outcome when matching features, computing non-uniform sampling strategies for speeding up sample-and-consensus model estimations, among others. I have created algorithms that use the statistical theory of extreme values to improve decision making processes in several computer vision and machine learning applications.


News

    Students

    Graduate Advisees

    Marcela Mera Trujillo (PhD)

    Alumni

    Eiad Kazkaz -- MS - Spring 2017. Now at LG.
    Xiang Gao -- MS - Summer 2018.
    Benjamin Smith (MS) -- MS Summer 2019. Now at Carnegie Robotics.

    Past Visiting Students

    Vishwajeet Narwal
    Richard Kralik

    Publications

    PatchMatch-based Neighborhood Consensus for Semantic Correspondence

    Jae Y. Lee, Joseph DeGol, Victor Fragoso, and Sudipta Sinha

    In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) 2021

    PDF Bibtex arXiv Video
    Generalized Pose-and-Scale Estimation using 4-Point Congruence Constraints

    V. Fragoso and S. Sinha

    In Proc. of the IEEE International Conf. on 3D Vision (3DV), Fukuoka, JPN, 2020

    PDF Bibtex arXiv Video Code
    Efficient Scene Compression for Visual-based Localization

    M. Mera-Trujillo, B. Smith, V. Fragoso

    In Proc. of the IEEE International Conf. on 3D Vision (3DV), Fukuoka, JPN, 2020

    PDF Bibtex arXiv Video
    BLT: Balancing Long-Tailed Datasets with Adversarially-Perturbed Images

    J. Kozerawski, V. Fragoso, N. Karianakis, G. Mittal, M. Turk, M. Chen

    In Proc. of the Asian Conf. on Computer Vision (ACCV), Kyoto, JPN, 2020

    PDF Bibtex arXiv Video
    Generalized Pose-and-Scale Estimation Given Scale and Gravity Priors

    V. Fragoso, J. De Gol, G. Hua

    In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR), Seattle, USA, 2020

    PDF Bibtex arXiv Code Video
    Hyper-STAR: Task-Aware Hyperparameters for Deep Networks

    G. Mittal, C. Liu, N. Karianakis, V. Fragoso, M. Chen, Y. Fu

    In Proc. of the IEEE/ Conf. on Computer Vision and Pattern Recognition (CVPR), Seattle, USA, 2020

    PDF Bibtex Video
    GraphMatch: Efficient Large-Scale Graph Construction for Structure from Motion

    Q. Cui, V. Fragoso, C. Sweeney, P. Sen

    In Proc. of the IEEE International Conference on 3D Vision (3DV), Quingdao, China, 2017

    PDF Bibtex Video
    ANSAC: Adaptive Non-minimal Sample And Consensus

    V. Fragoso, C. Sweeney, P. Sen, M. Turk

    In Proc. of the British Machine Vision Conference (BMVC), Imperial College London, London, England, 2017.

    PDF Bibtex
    Large Scale SfM with the Distributed Camera Model

    C. Sweeney, V. Fragoso, T. Hollerer, M. Turk

    In Proc. of the IEEE International Conference on 3D Vision (3DV), Stanford Univ., CA, USA, 2016.

    PDF (ArXiV) Bibtex
    One-class Slab Support Vector Machine

    V. Fragoso, W. Scheirer, J. Hespanha, M. Turk

    In Proc. of the IAPR International Conference on Pattern Recognition (ICPR), Cancun, QRO, Mexico, 2016.

    PDF (ArXiV) Bibtex
    Eye-CU: Sleep Pose Classification for Healthcare using Multimodal Multiview Data

    C. Torres, V. Fragoso, S. Hammond, J. Fried, B.S. Manjunath

    In Proc. of IEEE Winter Applications in Computer Vision (WACV), Lake Placid, NY, USA, 2016.

    PDF (IEEE Xplore) PDF (ArXiV) Bibtex
    Computer Vision for Mobile Augmented Reality

    M. Turk, V. Fragoso

    In Mobile Cloud Visual Media Computing: From Interaction to Service, Springer International Publishing, 2015

    PDF Bibtex
    gDLS: A Scalable Solution to the Generalized Pose and Scale Problem

    C. Sweeney, V. Fragoso, T. Höllerer, and M. Turk

    In Proc. of European Conference in Computer Vision (ECCV), Zürich, Switzerland, 2014.

    PDF Bibtex Code
    Cascade of Box (CABOX) Filters for Optimal Scale Space Approximation

    V. Fragoso, G. Srivastava, A. Nagar, Z. Li, K. Park, M. Turk

    In Proc. IEEE Conf. on Computer Vision and Pattern Recognitioin Workshops (CVPRW), Columbus, OH, USA, 2014.

    PDF Bibtex
    EVSAC: Accelerating Hypotheses Generation by Modeling Matching Scores with Extreme Value Theory

    Victor Fragoso, Pradeep Sen, Sergio Rodriguez, and Matthew Turk.

    In Proc. of IEEE International Conference on Computer Vision (ICCV), Sydney, Australia, 2013.

    PDF Bibtex Theia's implementation
    SWIGS: A Swift Guided Sampling Method

    Victor Fragoso and Matthew Turk

    In Proc. of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Portland, Oregon, USA, 2013.

    PDF Bibtex Theia's implementation
    Locating Binary Features for Keypoint Recognition using Noncooperative Games

    Victor Fragoso, Matthew Turk, and Joao Hespanha

    In Proc. IEEE International Conference on Image Processing (ICIP). Orlando, Florida, USA, 2012.

    Student Paper Award Finalist

    PDF Bibtex Code
    Automatic text detection for mobile augmented reality translation

    Marc Petter, Victor Fragoso, Matthew Turk, and Charles Baur

    In Proc. IEEE International Conference on Computer Vision (ICCV Workshops). Barcelona, Spain, 2011.

    PDF Bibtex Video
    TranslatAR: A Mobile Augmented Reality Translator

    Victor Fragoso, Steffen Gauglitz, Shane Zamora, Jim Kleban, and Matthew Turk

    In Proc. IEEE Workshop on Applications in Computer Vision (WACV). Kona, Hawaii, USA, 2011.

    PDF Bibtex Video

    Pre-prints

    Patch Correspondences for Interpreting Pixel-level CNNs

    V. Fragoso, C. Liu, A. Bansal, D. Ramanan

    PDF (ArXiV)

    Teaching

    West Virginia University

    Fall 2016 :: CS 350 - Computer Systems Concepts
    Website
    Fall 2016 :: CS 470 - Introduction to Computer Graphics
    Website
    Spring 2016, Fall 2017 :: CS 591B - 3D Computer Vision
    Website

    University of California, Santa Barbara

    Spring 2015 :: CS/ECE 281B - Advanced Topics in Computer Vision

    Software

    Mean Shift Clustering

    A threaded C++11 implementation using Eigen.

    Code
    Optimo: A C++ header library implementing several optimization algorithms.

    A C++11 implementation of several optimization algorithms.

    Code Documentation
    Statx: A C++ library implementing several MLE estimators with focus on Extreme Value Distributions.

    A C++11 implementation of several MLE estimators for several Extreme Value Distributions.

    Code

    Copyright © 2021 - Victor Fragoso