Selected work / Case study
Tono4
Learn Chinese vocabulary through the camera.
VerifyingOverview
An iOS app for learning Chinese vocabulary from camera input, built with SwiftUI, ARKit, and Core ML YOLO.
- SwiftUI
- ARKit
- Core ML
- YOLO
The problem
Tono4 explores learning Chinese vocabulary from objects around the learner. The intended flow connects camera recognition to saved words, pronunciation practice, and quizzes.
My contribution
I built an iOS vocabulary-learning prototype with camera exploration and practice modes. The README credits CoreML-in-ARKit as its foundation. These are source-level capabilities; a running app has not been verified.
How it's built
SwiftUI provides the interface, with ARKit, SceneKit, Vision, and bundled Core ML YOLO models for camera recognition. Core Data stores word collections. Audio and speech integrations support the intended pronunciation workflow. Model conversion utilities are separate Python scripts.
What's verified
Simulator build not verified.
As of Eyrie's October 29, 2026 check, the simulator build was not verified. No device behavior, screenshots, or demo is claimed.
What remains unverified
The simulator build failed during Core ML compilation in the evaluation environment, and simulator services were unavailable. No tests ran. Installation, recognition, pronunciation assessment, and review scheduling remain unverified. AR behavior needs a physical device check; no screenshots or demo are presented.