About
An independent student project
EmotionLens is an independent student research and learning project by Peilun Li. Through a literature review and interactive demonstrations, it explores how facial expression recognition works, where it breaks down, and what it does to the people it is pointed at.
Project scope
What this site is
- Science communication and research presentation
- Teaching simulations with visible, fixed rules
- Bilingual explanation of methods, evaluation and ethics
What it does not do
- Real emotion recognition of any person
- Medical, psychological or clinical diagnosis
- Camera access, photo storage, identity/age/gender inference, or any face history
How this site handles data
The teaching simulation runs in your browser from fixed rules written into the page. The Photo Expression Lab is the one place where an image is analysed: a photo you choose is decoded and classified by a pre-trained model that runs entirely inside your browser. The photo is never uploaded, never stored in a database or local storage, never logged, and no photo history is kept — clearing it or closing the page removes it. Only the model files are downloaded (and may be cached by your browser). The site never requests camera access and includes no analytics or tracking scripts.
Your language and theme preferences are stored in your browser's local storage so the site remembers them on your next visit. Clearing your browser data removes them. The experiment report you can download is generated locally and is not sent anywhere.
One honest caveat: this site is served by a third-party hosting platform, which — like any web host — may keep its own request logs. No absolute privacy guarantee can be made on that platform's behalf.