- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi,
In human pose estimation HTML pages its says that network is trained and evaluted on Coco dataset. But network outputs 18 keypoints instead of Coco (17 joints).
Is there any aditional information about joints and accuracy validation ?
Thanks,
Vurc.
Link Copied
2 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Please, refer to this description: https://github.com/opencv/openvino_training_extensions/tree/develop/pytorch_toolkit/human_pose_estimation#output-format
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you for the reply.
But still i dont get reported accuracy.
Thanks,
Vurc.

Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page