Yoshiteru, K.

Unfortunatelly, no. I'm not familiar with tflite-python-API.

Hi, It requires OpenGLES Shader code which runs on GPU. currently, shader code for YUYV format is ready as below. I think you can easily modify it to support UYVY....

I've implemented to support UYVY color format. Could you try **gl2blazeface** ? (Since I don't have UYVY camera, I can't check it.)

At first, 2.2.1. build TensorFlow Lite library on _**Host PC**_. And then, 2.2.2. copy Tensorflow Lite libraries to _**target Jetson / Raspi.**_

What is `BlazePose on Windowd 10` ? Which app did you use ?

Did you build the tensorflow library following the below procedure described in README ? - 32bit: https://github.com/terryky/tflite_gles_app#22-build-for-aarch64-linux-jetson-nano-raspberry-pi or - 64bit: https://github.com/terryky/tflite_gles_app#23-build-for-armv7l-linux-raspberry-pi

USB camera can be used without any source code modification. To use a mp4 video, follow the below instructions. https://github.com/terryky/tflite_gles_app#32-recorded-video-file

Yes, you are right. Unfortunately, some quantized models included in this repository may be low accuracy. I think there are several techniques to improve accuracy, but I am not familiar...

Hi @ragavendranbala, I'm not familier with **Tensorflow Lite python API**, but It's easy to implement custom Op in **C++ API**. Please refer follows: - https://www.tensorflow.org/lite/guide/inference#write_a_custom_operator - https://github.com/terryky/tflite_gles_app/blob/master/gl2hair_segmentation/tflite_hair_segmentation.cpp#L25-L26

Tensorflow Lite GPU Delegate supports following custom operations. If your implementation enables GPU Delegate, the custom ops run on GPU. https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/delegates/gpu/common/model_builder.cc#L2715-L2753