Projects

Here is a list of some of my favorite projects!

Anime on Ice

Present

Diane and Sabina are on a mission to create anime adaptations on ice! Our goal is to translate the meaningful messages we get from anime into art that skating fans, including those who are not familiar with anime, can ponder and enjoy. Check out some of our past figure skating routines here!

Present

Discord Soundboard Bot

February 2022

Developed a discord bot that enables users to play fun sound effects during voice calls. The bot includes an interactive soundboard that has multiple built-in category and sound effect options. Includes emoji-based buttons for a more intuitive user experience, as well as real-time threads to track user button presses.

February 2022

Tracking Dream SMP's Twitch Viewership Movements

May 2021

Ever wondered how Twitch viewers move between streamers, depending on who goes live? How do streamers' individual streaming habits affect the overall streaming landscape? How does Twitch viewership evolve through time? To answer these questions, I developed an interactive visualization of viewership movements between members of Dream SMP's Minecraft Twitch community. Final Project for 6.859 Interactive Data Visualization.

May 2021

Automatic Manga Colorization

May 2021

This project evaluates how GANs perform in colorizing manga by comparing the results from two models, Pix2Pix and CycleGAN. Both networks were trained on 2700 (black-and-white or grayscaled) and (colored) manga pages, and then tested with both black-and-white and grayscaled images, using the popular manga JoJo's Bizarre Adventure Diamond Is Unbreakable. See if you can differentiate between the real vs generated images! Final project for 6.869 Advances in Computer Vision.

May 2021

Visualizing Ludwig's 31‑Day Subathon

April 2021

On March 14, 2021, Twitch streamer, Ludwig Ahgren, started a subathon for which every new subscriber would add 10 seconds to the length of his stream. Due to overwhelming support from the Twitch community, the stream ended up lasting 31 days. During this time, Ludwig gained over 960k new followers and 250k new subscribers, enabling him to ultimately surpass Ninja's record for "Most Concurrent Twitch Subscribers" by hitting 280k+ subscribers at peak. I created an interactive visualization for viewers interested in learning more about activity statistics and stream highlights during this record-breaking event. Midterm project for 6.859 Interactive Data Visualization.

April 2021

Photochromic 3D Printing

March 2020 - May 2021

For my master's thesis, I built an integrated DLP 3D Printer that reduces the number of fabrication steps necessary to print single-material, multi-colored 3D objects using photochromic dyes. Research conducted under MIT CSAIL's HCI Engineering Group.

March 2020 - May 2021

Braille E‑Learner

March 2020 - May 2020

Braille E-Learner is a tactile, auditory, and visual multimodal learning system that teaches users Braille through an interactive feedback system. This system enables users to learn Braille by themselves without the need for an actual instructor on the side. Final project for 6.835 Intelligent Multimodal User Interfaces.

March 2020 - May 2020

Mobile Autonomous Systems

January 2020

Designed and built a robot to autonomously collect, sort, and score red and green cylinders into their respective goals. Our robot got 3rd place in the final competition! Final project for 6.146 Mobile Autonomous Systems Laboratory.

January 2020

Racecar Middle School

June 2019 - January 2020

Developed and taught a month-long summer curriculum that introduces students to the basics of programming, computer vision, and robotics for autonomous vehicles. I taught the pilot Summer and Fall school-year programs, several outreach events, as well as the Winter educator's training session for local teachers looking to incorporate more hands-on learning into their classes. This curriculum was developed as part of the MIT BeaverWorks Summer Institute, which aims to develop project-based courses to improve engineering education worldwide.

June 2019 - January 2020

Adaptive Accessories

September 2018 - December 2018

Designed and built adaptive earrings that changes shape according to user activity level and environmental stimulus. Uses flexinol wires as actuators, and pulse sensors and photoresistors as external sensors. Final project for 6.810 Engineering Interactive Technologies.

September 2018 - December 2018

Deep Learning Practicum

September 2018 - December 2018

Explored fundamental deep learning techniques via interactive user interfaces and practical application examples. Final project involved designing a deep learning network for predictive protein folding. Course assignment website for 6.S198 Deep Learning Practicum.

September 2018 - December 2018

Multi-touch Music Tiles

September 2018 - October 2018

A multi-touch pad music player that uses computer vision-based gesture recognition to trigger musical sound effects. Includes two different sound styles and eight different instruments to play from. Midterm project for 6.810 Engineering Interactive Technologies.

September 2018 - October 2018

Autonomous RACECAR

February 2018 - May 2018

Designed and implemented algorithms to make mini-racecars fully autonomous. Labs provided a comprehensive overview of mobile robotics and autonomous vehicles (ie. control and estimation theory, computational perception, computer vision, and motion planning). Project website for 6.141 Robotics Science and Systems.

February 2018 - May 2018

Class Projects

September 2016 - December 2020

A selection of additional interesting class projects that were super fun to implement! It was through some of these courses that I discovered my passion for developing interactive technologies. Documented here for mostly nostalgia purposes :)

September 2016 - December 2020

Map of Andorra's TripAdvisor Reviews

January 2016

Using geolocation, ratings, topic and sentiment breakdown data, I designed an interactive map that highlights specific business locations in Andorra based on selected user preferences. Features include search autocomplete, nationality filters, business type filters, business overviews, and heatmap/sentiment toggles. For example, try visualizing the token, "food". Research conducted for the Andorra Living Lab project, developed under the City Science group in the MIT Media Lab.

January 2016