Experience
-
-
-
-
-
-
-
-
-
-
-
Education
-
Stanford University
Activities and Societies: TreeHacks, Director of Bootcamp, a 10wk startup crash course. CS GPA: 4.0
Transferred to Stanford as one of five university transfer students.
-
Activities and Societies: Center for Artificial Intelligence in Society, Scope, Lavalab, Association for Computing Machinery, Society of Women Engineers, Applied Statistics Club
USC Full-Tuition Trustee Scholarship, USC Viterbi Fellowship and Scholars Award (top 2% of USC Class of 2021).
-
Activities and Societies: Founder of Club Code.
Led grassroots effort to establish the Whittier High School District's first engineering program by creating a partnership with Cal Poly Pomona. Over 10,000 students took the new STEM courses beginning in 2017. Featured in several national news articles and educational podcasts for my efforts.
Volunteer Experience
-
VAST Mentor and Teacher
University of Southern California
- 3 years 1 month
Science and Technology
Invited by Dr. Katie Mills from the University of Southern California to be the only high school student (along with undergraduate and master's level USC students) selected to tutor and mentor magnet school students in coding and robotics on the USC campus throughout the current school year.
Publications
-
Deep CNN Frame Interpolation with Lessons Learned from Natural Language Processing
arXiv
See publicationA major area of growth within deep learning has been the study and implementation of convolutional neural networks. The general explanation within the deep learning community of the robustness of convolutional neural networks (CNNs) within image recognition rests upon the idea that CNNs are able to extract localized features. However, recent developments in fields such as Natural Language Processing are demonstrating that this paradigm may be incorrect. In this paper, we analyze the current…
A major area of growth within deep learning has been the study and implementation of convolutional neural networks. The general explanation within the deep learning community of the robustness of convolutional neural networks (CNNs) within image recognition rests upon the idea that CNNs are able to extract localized features. However, recent developments in fields such as Natural Language Processing are demonstrating that this paradigm may be incorrect. In this paper, we analyze the current state of the field concerning CNN's and present a hypothesis that provides a novel explanation for the robustness of CNN models. From there, we demonstrate the effectiveness of our approach by presenting novel deep CNN frame interpolation architecture that is comparable to the state of the art interpolation models with a fraction of the complexity.
Projects
-
MatchSC
-
See projectEmploying state of the art AI techniques and well-researched psychological profiles, MatchSC quantified compatibility to pair USC students with the soulmate (platonic or romantic!) they never knew they haFor this project, I generated matches using graphs & machine learning, leveraging Instagram data to match students with peers based on position within the USC social graph and personality/values.
We got 4k+ USC sign ups within 24 hours from word of mouth alone, matching over 5.2k+…Employing state of the art AI techniques and well-researched psychological profiles, MatchSC quantified compatibility to pair USC students with the soulmate (platonic or romantic!) they never knew they haFor this project, I generated matches using graphs & machine learning, leveraging Instagram data to match students with peers based on position within the USC social graph and personality/values.
We got 4k+ USC sign ups within 24 hours from word of mouth alone, matching over 5.2k+ students a few days later. 96% of match recipients wanted a second match. Featured in a variety of news articles. -
Aper.io
-
See projectGiven any low quality video, Aper.io utilizes a series of Deep Convolution Neural Network Style Transfer Models which, given a series of frames, is able to predict smooth interpolation images. Due to the fact that this application is a largely unexplored space, we trained our predictive style transfer models from scratch. We did our prototyping on our own local machines, and once we had a solid foundation, we moved all our training onto Google Collaborate, which was key in meeting the deadline.
-
Seed
-
See projectSeed helps budding professionals keep up with their growing networks.
Used Vue, Firebase, HTML, CSS, and JavaScript to develop a relationship management tool that organizes contact information and tracks digital interactions. -
Phrase Party
-
See projectIf you Google "Markie Maraya Wagner", you will see that in the summer of 2012, I created and published an Android phone app called "Phrase Party." It was free, and my app received 40,000 downloads.
-
First Jobs
-
"First Jobs" is an app that finds entry-level jobs for high school and college students. Worked with the Director of the Boyle Heights Technology Center to pilot the app.
Honors & Awards
-
2nd Place @ LA Hacks, "Build Something People Want" Grant
UCLA
Developed a Machine Learning Algorithm that improves video streaming quality by predicting intermediate frames. Potential applications include video compression, live streaming, and simply improving old videos. Also winner of 1517's "Build Something People Want" Grant.
Devpost: https://devpost.com/software/aper-io
People also viewed
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More