asus4/tf-lite-unity-sample
A Unity plugin enabling TensorFlow Lite model inference across mobile and desktop platforms.

Velocity · 7d
+0.4
★ / day
Trend
→steady
star history
This repository ports TensorFlow Lite examples to the Unity game engine, providing pre-built libraries for running ML models on iOS, Android, macOS, Windows, and Linux. It includes implementations for tasks such as object detection (EfficientDet), pose estimation (MoveNet, BlazePose), face mesh, hand tracking, and style transfer. The project bundles GPU, NNAPI, and Metal delegates for hardware-accelerated inference alongside CPU fallback.