Inference using TensorFlowLiteSwift on iOS

[tucan9389]

Hi, I'm the maintainer of tucan9389/PoseEstimation-CoreML repo and I would like to share PoseEstimation-TFLiteSwift repo.

Recently, as TFLite support GPU accelerator using metal internally, I made PoseEstimation-TFLiteSwift repo, which inference various pose estimation models using TensorFlowLiteSwift library, for comparing with Core ML performance and library interface.

Providing the baseline code for pose estimation model and measure various pose model is the main purpose of the repo, which is similar to PoseEstimation-CoreML. And the repo support PoseEstimationForMobile models and google's PoseNet model

maybe support OpenPose tflite mode soon

If you want to see roadmap, please check the tucan9389/PoseEstimation-TFLiteSwift/issues.
Thanks for nice sample models 👍