Documentation on using 403_trt_pose_hand on non trt systems?

[avatarrokuaskingquestions]

Issue Type

Documentation Feature Request

OS

Ubuntu

OS architecture

x86_64

Programming Language

Python

Framework

ONNX

Model name and Weights/Checkpoints URL

trt_pose_hand, https://github.com/PINTO0309/PINTO_model_zoo/tree/main/403_trt_pose_hand

Description

Can run the trt_hand_pose models on any Ubuntu system but functions required to parse the information are all in the C++ layer of trt_pose (such as find_peaks). Have you heard of anyone reimplementing this code and if they'd be willing for you to link to it? Or if the functions are actually just basic enough that they're made in other libraries a pointer would help people to use the models in other setups.

Relevant Log Output

No response

URL or source code for simple inference testing code

No response

[PINTO0309]

I have little interest in NVIDIA's implementation, as I find processing at keypoints to be an unrealistic idea for real world use.

I think their implementation is hardly an implementation of DeepLearning because it is a very inefficient implementation that uses a lot of symbolic loops.

You might want to ask ChatGPT or something.

[avatarrokuaskingquestions]

Thanks for the quick response! Will keep that in mind...