Brent Lee

Electrical Engineering and Computer Science @ University of California, Berkeley

I'm a college student with a passion for computer engineering.
I'm highly interested in circuit boards, integrated circuits, embedded systems, machine learning, and computer vision.

I'm currently enrolled at UC Berkeley pursuing a Bachelor's Degree in Electrical Engineering and Computer Science.

I'm fascinated by the capabilities of artificial intelligence and want to explore AI as I pursue my interests in computer engineering.

See my

Skills Portfolio
Internship Resume

My Portfolio

Starflower Avionics Board

Avionics board built for the STAR onboarding project for our rocket Starflower (slide 4). The board uses an ESP32-S3-WROOM-1-N8 and features a working barometer and IMU. Code for interfacing with flash memory is W.I.P. Telemetry recording on ESP32 flash begins when IMU detects sufficient acceleration (takeoff!).

STM32 Mechanical Keyboard

I am currently working on building an STM32-based mechanical keyboard. The final version will have an I2C mini-OLED screen, pair of digital encoder knobs, and fully addressable RGB LED lighting. I eventually plan to have an aluminum case machined to house the PCB. Instead of KiCad, I'm using Altium Designer for schematic and layout. I've had the proof-of-concept macro-pad manufactured and have finished hand-soldering (pictured left), and have just started bringup.

Arduino Mechanical Keyboard

For this project, I designed and built my own Arduino mechanical keyboard. Using KiCAD, I created the schematic and designed the PCB. After sending it in to be manufactured, I used a combination of LibreCAD and Fusion 360 to design a keyboard case.

Video Github

AimLab Auto-Aim

Using computer vision to technically create an "aim-bot" for an aim training game. I used OpenCV and contours to detect targets. After compensating for a first-person-view, the mouse moves to and clicks on that target.

Github

TensorFlow Object Detection Assisted Annotation

In the earlier stages of my Auto-Aim project, I planned to use object detection to locate targets. Although that proved inefficient, in the process, I wrote a program to help me annotate the training data. The program uses a partially trained model to write annotation boxes to an xml file.