search

open-mmlab / mmpose

OpenMMLab Pose Estimation Toolbox and Benchmark.
https://mmpose.readthedocs.io/en/latest/
Apache License 2.0
5.82k stars 1.24k forks source link

rtmpose评估问题 #2078

Closed xiaochenwise closed 1 year ago

xiaochenwise commented 1 year ago

If I use my own data to train rtmpose, how should I evaluate it on the coco data set? I didn't find a module for evaluation, but there was this module in the previous version of mmpose.

mm-assistant[bot] commented 1 year ago

We recommend using English or English & Chinese for issues so that we could have broader discussion.

Tau-J commented 1 year ago

Thanks for using MMPose. You just need to set the dataset and val_dataloader to be coco data. Take this config as an example, the training and validation are all on COCO, you can modify the training part into your own datasets, and keep the val part as it was.

xiaochenwise commented 1 year ago

Thanks for using MMPose. You just need to set the dataset and val_dataloader to be coco data. Take this config as an example, the training and validation are all on COCO, you can modify the training part into your own datasets, and keep the val part as it was.

Hello, now that the model training is finished, I want to test the accuracy of COCO val2017 with detector having human AP of 56.4, but the command evaluated by the previous version will give an error

Tau-J commented 1 year ago

Hi @xiaochenwise , MMPose 1.x is incompatible with previous version, so you need to follow new docs to test. If you meet any error, you can print it in this issue.