Albert Han

IDEA Hacks is UCLA's annual hardware hackathon, co-hosted by IEEE at UCLA and Theta Tau at UCLA. This year, I had the pleasure of being one of three team captains in charge of the event. Due to the coronavirus pandemic, our plans of having the hackathon be hybrid had to be shifted to it being completely remote.

Despite the setbacks, we were able to successfully plan and execute the hackathon by having parts shipped to UCLA and having them packaged by a few volunteers. The parts were then shipped to hackers (or were picked up for students living near UCLA), who were then able to hack from home. In addition, we used Discord for annoucnements and general chatter, as well as Devpost for the projects.

Check out this year's IDEA Hacks website here and check out the projects at this year's devpost here!

In Winter 2021, I joined Daily Bruin as a DevOps Intern to work on the backend of the inventory website, designed to handle the checkout and checkin system at Daily Bruin. The service is designed to be used by hundreds of student organization leaders and managers.

Throughout the project, I worked extensively on the backend of the project, making several critical decisions about the design and structure of the database and its corresponding functions. I had to work with multiple levels of users, as well as a large list of equipment. Towards the end of spring quarter, I also worked on several pages of the actual site.

Languages/Technologies Used: JavaScript, React.JS, Firebase

In summer of 2020, I was a research assistant at the Center of Vision, Cognition, Learning, and Autonomy at UCLA, working under Tengyu Liu. Over the summer, I worked with combining SMPL (body) and MANO (hand) models, utilizing Python libraries such as Tensorflow and Numpy to predict the entire body's mesh using a list of vertices.

The image above shows a part of the process. The image on the left shows an example of the first part of the process, which was transforming the hand to the right position and angle given the vertices of the body using SMPL. Then, I would figure out the individual transformations for each joint in the hand model, as seen on the right. Finally, I would connect the two, and run the model to get the final surface of the body and hand.

Languages/Technologies Used: Python, PyTorch, TensorFlow

Two part, bluetooth connected device. Designed for weary travelers, the motion sensor is placed inside the luggage, while the user holds the second piece. If the motion sensor senses movement (the sensor is detected by a gyroscope), the second piece will vibrate uncontrollably (thorugh a motor) until the user places their fingerprint on the sensor.

Technologies Used: Arduino, Gyroscope, Bluetooth, Fingerprint Module

Interactive escape room game where the player must find clues hidden throughout the room and escape. In addition to using the WebGL/tiny graphics library, we also heavily utilized Adobe Illustrator and Blender to create objects and designs that would immersify the player into the room.

Through this project, we decided to be extra detailed with the design of the environment, and add plenty of designs and custom made objects. Check out the repository of the project here!

Languages and Technologies Used: JavaScript, WebGL Blender, Adobe Illustrator

For our capstone class, my team and I had to create an application using C++ and the boost library, utilizing NginxConfig to declare handlers and creating a complex request handler that responded to various requests. Our final project, Coding Homies Community, was a site where users could create topics and write posts underneath the topics, and other users could add on by adding posts to each topic.

The project included a robust API with the paths defined in a NginxConfig file, which was interpreted by a parser. The corresponding request parser would handle the request appropriately, sending a boost::beast::http::response object. In addition to source code, our team wrote thorough unit and integration tests. Our final project was deployed onto Google Cloud Platform by utilizing Docker containers.

Languages and Technologies Used: C++, SQLite, CMake, Docker, Google Cloud Platform, NginxConfig

For my summer personal project, I built a complex flashcard site where the use can create a set of terms, practice by flipping through flashcards, and then completing a test. The project was built with React.JS, Node.JS, Firebase, and MongoDB.

The user can create a new set, where he/she will define the title and the terms and definitions of their items. The title, terms, and definitions can be edited if necessary.

The user can then practice the set using a online flashcard-like system. This can save paper and can help create a pleasant studying experience.

The user can finally test their knowledge by ensuring that they can connect a particular term to its correct definition. After taking the test, the user can view their results, find out their mistakes, and continue practicing or retake the test.

In addition to the mentioned features, benky uses authentification - thus, sets and test results are connected to an email. Authentification for this application is managed using Firebase.

Languages and Technologies Used: React.JS, Material UI, Node.JS, Express.JS, MongoDB, Firebase

As the publicity chair for the IEEE Student Branch at UCLA, I have produced various designs for the club and have led a team to create all the graphics for the club's annual hardware hackathon, IDEA Hacks.

In addition, I have done graphic design for other projects, namely Night in the Diner and this website! Here are a few of the designs that I am most proud of.