Women in Computer Vision

CVPR 2017 Workshop

Music education has proven academic and social benefits, but is costly and time intensive. This workshop contribution aims to hallucinate an instrument based on the marimba from the user’s actions. To this end, we present a novel dataset, current progress in generating pitched notes that temporally correspond to the action, velocity, and location of a strike, and future direction for the generation of music from video.

Workshop Paper

About Me

I am a second year Computer Science Ph.D. student at the Georgia Institute of Technology, advised by Professor James Hays. I am currently working on image to sketch generation and generative networks across the image and music domains. My research interests are in computer vision and deep learning, specifically, the use of generative networks for augmenting human creativity.

I was a research intern at MIT Lincoln Laboratory (person re-identification), Apple (RGB-D scene semantics), and a National Science Foundation Research Experience for Undergraduates at the University of Central Florida with Professor Mubarak Shah (Deep Tracking). I received my Bachelor in Science in Computer Science from the George Washington University with a minor in Music.

Outside of research, I enjoy playing violin or viola in chamber groups, going to concerts, and wandering around the High. If I'm not in the lab, I'm probably chasing my rescue pup Benji around the park!

Research

Resume

 

 

GT Computational Perception Laboratory Current

Generating human-like sketches from photos, a task that requires generative semantic notions of objects to be disentangled from direct spatial alignments. While previous work has shown promise in generating segmentations from images, emojis from faces, and even photos from colored sketch strokes, the method used for these tasks relied on the training data for the neural networks to be spatially aligned. As we know, human sketches are not so well aligned with the sketches that are drawn. This exploration will highlight ways in which to make image generation more spatially invariant, as well as better methods to encode sketch information.


Image to Music Generation

GT Center for Music Technology Current

Generating music for non-musicians: when a user taps different items, the video is analyzed and visual features are correlated to music features that were learned during training on a newly collected marimba audiovisual dataset.


MIT Lincoln Laboratory

Summer Research Program Intern June 2016 - August 2016

G104: Intelligence & Decision Technologies Group, Leveraged multi-network deep learning for person re- identification across multiple camera views. Conducted ablation study to better intuit the effects of parameter training, hard negative mining, and feature vs. neighborhood feature difference on Siamese networks.


Stanford U.S.-Russia Forum

Technology & Entrepreneurship Working Group May 2015 - April 2016

Analyzing American and Russian models of technological innovation such as Silicon Valley and Skolkovo. Using econometrics and qualitative social studies to understand the variables driving entrepreneurship and innovation in naturally and artificially formed clusters of innovation. Creating a paired comparison of various clusters of innovation to suggest ideal models for encouraging innovation in the US, Russia, and in both naturally and artificially formed innovation centers.


Apple

Software Engineering Intern May 2015 - August 2015

Researched semantic understanding of scenes using RGB-D cameras to create a feature that is currently being expanded.
iContest Finalist: Participated in an Apple hackathon, presented a novel feature to a jury of managers.


National Science Foundation Research Experience for Undergraduates

Research Fellow May 2014 - August 2014

Advisors: Dr. Mubarak Shah & Dr. Niels de la Vitoria Lobo
University of Central Florida, Center for Research in Computer Vision
Created Deep Tracking, which improved state of the art tracking performance by over 12% with the application of deep learning techniques to the fundamental challenge of tracking. Introduced motion and appearance dual-flow convolutional neural networks to tracking. Created a novel approach to scale change that is feature-dependent, rather than scale-dependent. The technical report can be viewed on Arxiv.


GW College of Medicine

Genetics & Breast Cancer Research 2011 - 2012

Advisors: Dr. Sidney Fu
Analyzed the role of microRNA-298 regulating genetic expression in breast cancer with genetic sequencing.


Albert Einstein College of Medicine

Genetics & Neuroscience Research 2011 - 2012

Advisors: Dr. Zaven Kaprielian & Dr. Hannes Buelow
Isolated the rz-1 mutation, which is responsible for extreme dendritic abnormalities, in C. elegans strains. Used genetic sequencing to find the gene responsible for the mutation. Chronicled its effects on dendritic formation as a preliminary model for understanding Fragile X Syndrome.


New York University

Visual Cognitive Psychology Research 2010 - 2011

Advisors: Dr. Marisa Carrasco & Dr. David Carmel
Studied the effect of attention cues on memory within the phenomenon of Boundary Extension.

Merit

College of Computing Awards: 7001 Research Award

Georgia Institute of Technology College of Computing Fellowship 2017

Global Art Nabi Hackathon: 2nd Place

Won 2nd place in a hackathon focusing on art and artificial intelligence in Seoul, Korea. 2016

Presidential Scholar

Georgia Institute of Technology College of Computing Fellowship 2016 - Graduation

Dean's Scholar

The George Washington University School of Engineering & Applied Sciences Scholarship 2012 - 2016

University Honors Program

GW Program of Academic Merit 2012 - 2016

Association for Computing Machinery

Grace Hopper Celebration, Student Research Competition October 2015

George Ellowitz Memorial Award

Award for arts-related travel. 2015

Shenkman Development Grant

Award for travel to Russia for research. 2015

Facebook Grace Hopper Travel Grant

Travel scholarship from Facebook. One of 25 awarded internationally. 2015

Google Grace Hopper Travel Grant

Travel scholarship from Google. 2014

Teaching

Undergraduate Teaching Fellow

CS1010: Introduction to Computer Science Fall 2015

Assisted students in lecture and lab sessions and guided them through building their final robotics project. Introduced students to algorithmic concepts in computer science such as stacks and queues.

Service

A personal passion of mine is music—whether it was DJ-ing a radio station in undergrad, spending high school entrenched in multiple orchestras, or my current exploration into music generation research, I've tried to incorporate my love of music into my life. I'm happy to be volunteering with Artportunity Knocks indefinitely, going to different elementary schools around Atlanta every week with my violin to provide children with a fundamental love and understanding of classical music. For any requests, please feel free to contact me at myname [at] gatech [dot] edu.

Technical Skills

  • Python
  • Java
  • C++
  • C
  • Matlab
  • Tensorflow
  • Torch
  • Theano
  • OpenCV
  • Depth imaging: Tango, Structure, Kinect
  • Caffe

Elements

Text

This is bold and this is strong. This is italic and this is emphasized. This is superscript text and this is subscript text. This is underlined and this is code: for (;;) { ... }. Finally, this is a link.


Heading Level 2

Heading Level 3

Heading Level 4

Heading Level 5
Heading Level 6

Blockquote

Fringilla nisl. Donec accumsan interdum nisi, quis tincidunt felis sagittis eget tempus euismod. Vestibulum ante ipsum primis in faucibus vestibulum. Blandit adipiscing eu felis iaculis volutpat ac adipiscing accumsan faucibus. Vestibulum ante ipsum primis in faucibus lorem ipsum dolor sit amet nullam adipiscing eu felis.

Preformatted

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';

Lists

Unordered

  • Dolor pulvinar etiam.
  • Sagittis adipiscing.
  • Felis enim feugiat.

Alternate

  • Dolor pulvinar etiam.
  • Sagittis adipiscing.
  • Felis enim feugiat.

Ordered

  1. Dolor pulvinar etiam.
  2. Etiam vel felis viverra.
  3. Felis enim feugiat.
  4. Dolor pulvinar etiam.
  5. Etiam vel felis lorem.
  6. Felis enim et feugiat.

Icons

Actions

Table

Default

Name Description Price
Item One Ante turpis integer aliquet porttitor. 29.99
Item Two Vis ac commodo adipiscing arcu aliquet. 19.99
Item Three Morbi faucibus arcu accumsan lorem. 29.99
Item Four Vitae integer tempus condimentum. 19.99
Item Five Ante turpis integer aliquet porttitor. 29.99
100.00

Alternate

Name Description Price
Item One Ante turpis integer aliquet porttitor. 29.99
Item Two Vis ac commodo adipiscing arcu aliquet. 19.99
Item Three Morbi faucibus arcu accumsan lorem. 29.99
Item Four Vitae integer tempus condimentum. 19.99
Item Five Ante turpis integer aliquet porttitor. 29.99
100.00

Buttons

  • Disabled
  • Disabled

Form