Little Tracks, Big Heart

Unity3D, environmental activism

Little Tracks, Big Heart is an exploration of how video games can be designed to help educate and encourage activism from players, while maintaining the fun and rewarding aspects of gaming. Video games have rapidly become integrated in almost every part of society- from educational uses to therapy to career options, and have the potential for so much more. I'm personally very passionate about environmental awareness, and thought it would be an interesting endeavor to create a game that incorporates current issues into the gameplay, followed by an option for players to enact real change. To do so, I designed and created several levels using Unity 3D that address certain environmental challenges, and upon completion of each level the player is directed to a relevant cause on change.org. In this way players can become aware of issues, invested in them, and then active in the solution- whether it's signing a petition, sharing on social media, or making a donation to a cause.

Global Goals

Game dev, social justice

Global Goals is a project I collaborated on with Daria Schifrina for Venus Hacks 2021 which won Best Hack for Social Good. The project devpost can be accessed here, and we have a YouTube walkthrough for it as well. It's a browser-based game hosted here that encourages social activism through a community-oriented app that enables players to grow an oasis of animals by engaging in social justice work. To earn rewards, players can donate to non-profit organizations, share causes on social media, and sign petitions. Our project is hosted on Replit, uses Python, Flask, SQLite. All graphics were proudly designed and drawn by us!

Eyes Free Mobile App

Mobile app, accessibility

This research project focused on examining mobile services available to the visually impaired community, and culminated in an application + paper describing our research process, app development, and study findings. Our solution took shape as a centralized, eyes-free application to help blind or low-vision individuals accomplish many of their day-to-day needs. For dev/evaluation we used the Expo platform and React Native framework, using libraries and sensors compatible with both Android and iOS: the microphone, accelerometer, and camera. Our backend is set up with Google Cloud, leveraging Google Cloud APIs for object identification, speech-to-text, text recognition, and text-to-speech.

ALice In Wonderland Escape Room

VR, Game dev

I was inspired by all the challenges in Alice and Wonderland, so decided to create an escape-room type game, but in VR. The world was created using a-frame + models availble on turbosquid. The player sees through the perspective of Alice, and is prompted to follow a white rabbit to the base of a tree where you then fall down the rabbithole and are transported to wonderland. Except you're stuck in a room, surrounded by strange objects, and a locked door seems to be the only way out. You can find it here or watch a quick video of gameplay in the browser! *a-frame is supported best by firefox and takes a few minutes to load all assets*

TaskMapper

Task management, Game dev

TaskMapper aims to help students with managing both personal tasks and academic assignments by providing a platform to track and incentivize work completion with meditative exploration. The browser based platform/game allows users to optimally track their tasks through a consolidated list that includes assignments and exams automatically imported from Canvas, with the option to modify and add as needed. The task list will be paired with a meditative world such that the user can unlock more areas to explore by completing tasks, and the world will contain varying breathing and meditative exercises to allow users to stay focused and relaxed while working. The platform was built with a React framework alongside a-frame and javascript, and uses the Canvas API. This project is currently published here, and the codebase is located at our github repo. Feel free to check it out!

ML Teaching App

Machine Learning, Mobile app

This research project was the end product of an internship in the Micrsoft Research CHIL group to make machine teaching more accessible to everyone, not just to machine learning experts. Teaching for machine learning requires humans to put what they know into a computer, often a time-consuming task- what if a there was a way to help that human do it in small bites? We developed an app that allows users to easily create and train their own machine learning model to classify their data into categories to organize and personalize their information flow. After performing a competitive analysis of services already available, we designed and developed this mobile application to improve upon existing products while incorporating new services using Expo, React Native, and Typescript/C#. Here is a video showing app usage/interaction as well as an article published about the project by the Microsoft Research blog.

Music Gallery

GAN, algorithmic music

An immersive experience that plays around with generated music and art through many outlets. For pre-generated songs I used RunwayML to style transfer the photos using various models, and generated the jukebox tracks with pretrained models as well. The "generate song" page produces a custom track each time and uses tone.js, while the "create song" page allows a user to interact with a keyboard model to test and record their own song. Overall this was just a casual exploration of computer/ai-generated music that is definitely unpolished, but fun to play around with and fun to make! *a-frame is supported best by firefox and takes a few minutes to load all assets*