Hi, I'm Zihan

Computer Engineering student focused on computer vision, robotics, AI systems, and full-stack engineering.

ABOUT ME

IamZihan(Henry)Zhao,aComputerEngineeringstudentatUCSBfocusedoncomputervision,robotics,andAIsystems.Ibuildperceptionpipelinesthatconnecttrainedmodelstoreal-worldmachines,alongwithfull-stacktoolsthatturncomplexworkflowsintoreliableproducts.

Computer Vision

Robotics Systems

AI Engineering

Full-Stack Development

Dean's HonorsAI, Robotics & Computer Vision concentrationOn-device inference and real-world deployment

Experience

01June 2026 - September 2026

Shanghai Weiduo Technology Co., Ltd.

World Robot Contest Park Vision System

Computer Vision Software Engineering Intern / Shanghai, China

Engineered an accuracy-first YOLO and VLM vision system for autonomous park inspection in the World Robot Contest.

Architected a multi-stage inference pipeline combining YOLO detections with full-frame and cropped-image VLM review

Curated a 3,500+ image dataset across 14 classes and trained YOLO11m through a 200-epoch workflow

Deployed on NVIDIA Jetson AGX Thor with Docker, vLLM, and reusable Python/CLI integration

3,500+ images14 classes150+ tests2nd of 47 finalists

PythonPyTorchYOLO11mQwen2.5-VLvLLMDockerJetson AGX Thor

GitHub Repo

Official Competition Footage

WRC / Park Scenario

AGIBOT humanoid robot collecting a bag of litter from the park course

Litter Collection

AGIBOT humanoid robot carrying an item beside the competition waste station

Waste Classification

AGIBOT humanoid robot navigating the marked park course

Autonomous Navigation

AGIBOT humanoid robot identifying the waste station on the park course

Waste Recognition

02June 2025 - September 2025

D-Dao Technology

AI Subtitle Generation Platform

Software Engineering Intern / Cupertino, CA

Built an end-to-end AI platform that transforms raw video into punctuated, Chinese-translated subtitles and downloadable results.

Integrated FFmpeg, WhisperX, and Gemini for speech recognition, translation, and subtitle generation

Developed asynchronous Flask services for inference orchestration, task scheduling, and backend APIs

Optimized CUDA-enabled GPU inference on Linux for large-scale video processing

End-to-end AI pipelineGPU inferenceAsync processingProduction-ready web app

PythonFlaskWhisperXGeminiCUDAFFmpegSQLAlchemySQLite

GitHub Repo

Subtitle Processing Console

GPU Ready

Product Demonstration20 min demo

Processing PipelineFive-stage workflow

Stage 01

MP4 Upload

Stage 02

FFmpeg Audio

Stage 03

WhisperX Speech

Stage 04

Gemini Translation

Stage 05

Subtitle Render

Featured Projects

01June 2025 - Present

Full Stack / Team Leadership

UCSB CSSA Official Website

Lead Developer / Santa Barbara, CA

Led the design and launch of UCSB CSSA's production website, creating a scalable bilingual platform for a 2,000+ student community.

Built reusable pages and server-side APIs with the Next.js App Router

Enabled non-technical officers to publish events, announcements, and media through Google Sheets

Delivered bilingual content, dynamic routing, and responsive user experiences

5-person dev team2,000+ students5,000+ page viewsProduction platform

Next.jsReactTypeScriptTailwind CSSCloudflare R2VercelGoogle Sheets

Visit Live Website

ucsbcssa.com

Live

UCSB CSSA

Chinese Students & Scholars Association

EventsResourcesAbout

Connect · Culture · Community

A digital home for the UCSB community.

连接社区 · 庆祝文化 · 共同成长

BilingualEventsResources

UCSBCSSA

2,000+ Students5,000+ Views

Live Events

Dynamic updates

中

Bilingual

Community access

↗

Resources

Student guides

Content Delivery Architecture

No-code publishing

1Editorial

Google Sheets

2Application

Next.js APIs

3Media

Cloudflare R2

4Delivery

Vercel

02January 2025 - March 2025

Hardware / Embedded Systems

Morse Code Converter

Developer / Santa Barbara, CA

An Arduino-powered physical device that converts text into Morse code and writes it using a servo-actuated pen and stepper-motor paper feeder.

Programmed Morse translation logic and joystick-controlled text input

Integrated real-time LCD feedback for the complete input workflow

Calibrated servo pen angles and stepper timing for consistent writing

Text-to-Morse logicReal-time LCD feedbackElectromechanical output3D-printed frame

ArduinoC/C++Servo MotorStepper MotorLCDJoystickBreadboard Circuits3D-Printed Frame

Physical Prototype Demonstration

System Online

Electromechanical Signal Path

From text input to marks on paper.

Arduino / C++ / Motor Control

Joystick + LCD

Input and feedback

Arduino Logic

Translate and time

Servo Pen

Write dots and dashes

Stepper Feed

Advance the paper

Skills / Tech Stack

Discipline 01

Languages

PythonC++TypeScriptJavaJavaScriptSQLHTML/CSS

Discipline 02

AI / Computer Vision

PyTorchYOLO11mQwen2.5-VLUltralytics YOLOvLLMWhisperXGeminiCUDAAI Pipelines

Discipline 03

Web / Application

Next.jsReactFlaskTailwind CSSBootstrapSQLAlchemySQLiteREST APIsVercel

Discipline 04

Systems / Deployment

DockerNVIDIA JetsonArduinoGitLinuxCodexFFmpegVerilog HDLServo MotorsStepper MotorsBreadboard Circuits

Opportunities / Collaborations

Let's Connect

I'm interested in internship opportunities and technical collaborations at the intersection of computer vision, robotics, AI systems, and full-stack engineering.

Email->

henryzhao@ucsb.edu

LinkedIn->

linkedin.com/in/zihanz2028

GitHub->

github.com/zihanzzh

Santa Barbara, CAUC Santa BarbaraComputer Engineering