Selected ExperimentAR · Computer Vision

Filter Reality

Real-time AR filters and gesture-driven overlays — tracking fingers and faces to augment the camera feed and even control physical output based on hand position.

Finger tracking driving real-time overlays and physical LED output.

Snapshot

Role: Designer & developer
Type: Real-time AR / computer vision
Inputs: Hand & face tracking
Tools: OpenCV · MediaPipe · Python

Overview

Mapping gesture to expressive output

Filter Reality explores how expressive, low-latency interaction can come from a single camera. By tracking hand and face landmarks, it overlays real-time AR treatments and maps gestures to visible output — including driving physical LEDs from hand position — to test how legible and responsive camera-based interaction can feel.

Challenge

Map expressive gestures to clear, legible output without lag or jitter.

Outcome

Reliable real-time tracking with readable AR overlays and responsive physical output.

View on GitHub

See the featured case studies

View work →Contact me