[Feature] rtmpose onnx inference example

open-mmlab / mmdeploy

OpenMMLab Model Deployment Framework

https://mmdeploy.readthedocs.io/en/latest/

Apache License 2.0

2.74k stars 629 forks source link

[Feature] rtmpose onnx inference example #1954

Closed sun11 closed 1 year ago

sun11 commented 1 year ago

Motivation

Currently the document of rtmpose inference do not contain the part of postprocessing code, and the post processing is slightly different from topdown_heatmap method. Could someone provide a example for rtmpose onnx model postprocessing?

Related resources

No response

Additional context

No response

irexyc commented 1 year ago

Could you point out what is the difference of the postprocessing? I didn't ses there is any different about val_pipeline or test_pipeline compared to other top down methods like hrnet

sun11 commented 1 year ago

Update: I figured this out by myself finally https://github.com/open-mmlab/mmpose/blob/dev-1.x/mmpose/codecs/utils/post_processing.py#L41