ViTAE-Transformer / ViTPose

The official repo for [NeurIPS'22] "ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation" and [TPAMI'23] "ViTPose++: Vision Transformer for Generic Body Pose Estimation"
Apache License 2.0
1.39k stars 187 forks source link

How to Evaluate Using ViTPose and a Custom Dataset in COCO Format #153

Open bgcomYSTO opened 1 week ago

bgcomYSTO commented 1 week ago

Greetings, I am a fourth-year university student in Japan conducting research on 2D pose estimation. As part of my research, I plan to use the package you have provided to evaluate a custom dataset I created. Could you kindly provide guidance on how to evaluate a custom dataset using your package? Thank you for your time and assistance.

bgcomYSTO commented 1 week ago

The custom dataset is created in CVAT and exported in COCO format.