Ellie Zhou

Ellie Zhou

Hi, I’m Ellie.

I am an AI researcher, builder, and aspiring entrepeneur. In general, I am an observant and grounded thinker who searches for nuance and stories in the margins. This theme of looking beyond the surface drives my work.

In tech, I've researched how to use AI to make healthcare more accessible to rural populations, and investigated AI's hidden biases. In my community, I advocate for unhoused youth through Valley Water and work to bring diverse voices into CS through ParagonX and various self-started school organizations. I use photography to capture and highlight the fleeting moments that I find meaningful.

Portrait of Ellie Zhou

Research

In my opinion, computer vision has the potential to truly shape the world because it gives machines the ability to perceive and understand the physical world, enabling them to solve problems that were once impossible. Inspired by this, my research explores how AI and computer vision can uncover insights in fields like medicine. However, these systems can only improve lives if they are accurate and reliable for everyone.

As AI becomes increasingly embedded in society, I believe it must also be fair and trustworthy. My work ranges from improving heart-rate estimation across diverse skin tones to reducing background bias in action recognition models, with the goal of building AI systems that people can confidently rely on.

Three panels: a video frame of a person's face, a black-and-white skin mask, and the resulting segmented image showing only skin regions

Developing a Non-Contact Cardiac Monitoring System via Remote Photoplethysmography (rPPG)

  • Won 2nd place at the Synopsys Championship.
  • Developed a real-time, low-cost, contactless AI model to monitor heart health, addressing a lack of insight into heart health among underserved and rural populations.
  • Achieved state-of-the-art results, predicting heart rate to within a single beat per minute.

See my poster

Two failure cases: a violinist on a baseball field labeled playing baseball, and a woman dancing in a supermarket labeled shopping

Seeing Beyond the Scene: Analyzing and Mitigating Background Bias in Action Recognition

  • Published to the NeurIPS 2025 NextVid and SpaVLE workshops.
  • Won 1st place at the Synopsys Championship.
  • Conducted the first systemic analysis of background bias across all AI models. After discovering that all rely on background as a shortcut, I built novel architectures and an automated prompt-generation system to mitigate these biases, outperforming state-of-the-art action recognition models.

Read the paper See my poster

Personal Projects

Building technology is a way for me to help my community. Everything I build is centered around one question: how can I improve the lives of those around me? This has led me to build an in-page translator for friends learning English, a website to help my friend’s father’s small business reach more customers, and a gaze-tracking communication tool for my grandpa with ALS. In college and beyond, I hope to dedicate myself toward building intelligent human-centered systems that meaningfully improve people’s quality of life.

LinguaLens

  • An in-page reading aid for English learners: select any word or phrase to get an AI translation, a simple-English definition, an image, and pronunciation — without leaving the page you are reading.
  • Chrome MV3 extension, running on Gemini or Cerebras.

View on GitHub

The Alex's Plumbing and More homepage: a dark hero reading Bay Area Plumbing, Concrete and Fencing, with call and free-estimate buttons

Plumbing Website

  • A website for my friend’s dad, who runs a multi-trade business in the Bay Area — plumbing, concrete, and fencing.
  • Helped increase brand visibility and attract more customers

Visit the site

The communication board on screen: a three-by-three grid of phrase tiles reading Yes, No, Thank you, I'm in pain, I need help, Water please, I need a break, and Keyboard, with one tile filling orange as the dwell timer runs, and a small webcam preview in the corner marking the tracked pupil

GazeTracker

  • A communication board you operate with your eyes: look at a tile, hold for about a second, and the computer speaks for you.
  • Built with Python and OpenCV, on top of an open-source eye tracker.

View on GitHub

Service & Leadership

Valley Water Youth Commission - Vice Chair

I was elected to serve as Vice-Chair for Valley Water’s 21-member Youth Commission, representing Santa Clara County’s 2M water consumers. I mentor Youth Commissioners, coordinate quarterly meetings with agency executives aligned to district goals, and lead projects.

Hygiene Kit Project

With San Jose having one of the highest per capita youth homelessness rates, I led a group of Youth Commissioners to organize a hygiene kit drive for youth facing homelessness. We raised more than $5000 that funded 100 kits, and hand-delivered them to the Bill Wilson Center. Our project helped restore dignity to vulnerable youth across Santa Clara County.

River Cleanups

As part of the Data Accessibility Team, I hosted three river cleanups at Alviso Slough and Coyote Creek, removing more than 750 pounds of trash. I coordinated logistics, delivered pre-event briefings to volunteers, and facilitated a trash data collection activity to identify the most common pollutants and discuss ways to reduce them at the source.

In the Media

  • The hygiene kit drive, featured by Valley Water (@scvwd).
  • Coastal Cleanup Day 2025 with the Youth Commission, featured by Valley Water (@scvwd).

ParagonX Academy - President

Teaching and mentoring students is one of my greatest joys. For me, learning to code and build AI models was transformative because it made me feel like I could build anything. Through teaching USACO Bronze and Intro to Machine Learning classes at ParagonX for two years, I strive to help other students achieve that same agency and confidence. Now, as President, I drive our broader organizational vision — leading the executive team, expanding our global outreach efforts to democratize education, and scaling our curriculum. So far, we have impacted 3,500 students, taught more than 480 classes, and reached 24 countries.

Visual AI & Computer Vision Club - President

“AI will change the world, but who will change AI?” is a quote by Professor Olga Russakovsky that really resonated with me. At programming competitions and hackathons, I always noticed that it seemed to be dominated by the same narrow demographic. Echoing Professor Russakovsky's words, my goal with this club is to help guide more students into AI, bringing more diverse voices into the field. As President, I prepare and give engaging lectures, demystifying AI and showing them real-world applications, and host games and competitions.

Science Fair Initiative - Founder

Students at my school had immense scientific curiosity but were unsure how to conduct research. I founded an initiative to help them realize those aspirations. I ran outreach across campus and brought in more than twenty students — triple the previous year’s participation. I organize our weekly meetings and mentor students 1-1 as they develop original projects.

Wharton Investment Competition - Team Lead

Freshman year, I led a team of four girls with no prior investment experience to become one of the top ten Global Finalists out of more than 1,600 teams. I developed the investment strategy for our client, designed an innovative trading algorithm, and analyzed IT and Consumer Discretionary assets. As team leader, I managed project milestones, fostered a collaborative and positive environment, and kept our team morale high throughout the eight-month competition.

Writing & Photography

Writing short stories and photography are the ways that I unleash my creativity. My writing explores themes of familial relationships, often echoing my own experiences. In photography, I love being able to slow down and capture the little moments, as well as being creative through digital editing in Photoshop — check out my strawberry panda!

Short Stories

Photography

Dance & Artistic Swim

Dance and swimming give me immense joy. I love the adrenaline rush of performing on stage and in the water. Both have also taught me essential skills that have transferred to other areas of my life: teamwork, resilience, and the courage to always keep trying and growing despite there not being immediate results.

Awards

  • Y Combinator Startup School Attendee (2026) 5% acceptance rate, more than 30,000 applicants. Heard from founders like Jensen Huang, Sam Altman, Boris Cherny, Chelsea Finn, and Alexandr Wang. Received 25K in compute credits.
  • 1st Place, Synopsys Silicon Valley Science and Technology Championship (2026) Qualified to the California Science & Engineering Fair.
  • IEEE Technical Excellence Award (2026)
  • Honor Roll, USA North American AI Olympiad Round 1 (2026)
  • CUHSD Certificate of Recognition for STEM Excellence (2026)
  • First-author paper on background bias accepted to two NeurIPS workshops (2025) NextVid and SpaVLE.
  • Second Place, Synopsys Silicon Valley Science and Technology Championship (2025)
  • USA Junior Mathematical Olympiad (USAJMO) Qualifier (2024) Qualified for the American Invitational Mathematics Examination twice, in 2023 and 2024, and for Math Prize for Girls in 2024.
  • Global Finalist, Wharton Global High School Investment Competition (2024) Team Leader. Top 10 of more than 1,600 global teams — under 1%.
  • Presidential Volunteer Service Gold Award (2024)
  • Finalist, Principles of Business Management & Administration Roleplay and Testing, California DECA State Conference (2024) 8th of 114 participants
  • State Honor, MTAC Certificate of Merit Advanced Level (Level 10) Piano (2024)
  • USA Computing Olympiad (USACO) Gold (2023)
  • Gold Key, Scholastic Art and Writing Awards, Short Story (2023)
  • Honorable Mention, Synopsys Silicon Valley Science and Technology Championship (2023) California State Science & Engineering Fair qualifier.