Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Automatic annotation for landmark/pose prediction #5324

Closed
sangdv opened this issue Nov 19, 2022 · 4 comments · Fixed by #7033
Closed

Automatic annotation for landmark/pose prediction #5324

sangdv opened this issue Nov 19, 2022 · 4 comments · Fixed by #7033
Assignees
@bsekachev
Labels
models

Comments

@sangdv
Copy link

sangdv commented Nov 19, 2022

Hi, we want to integrate our landmark/pose prediction model into CVAT for automatically suggesting keypoints. But it seems that CVAT allows integrating only interactors, detectors and trackers, which are not suitable for the landmark prediction task. Could you please guide me on how to do that?
@nmanovic nmanovic added the models label Jan 3, 2023
@martin0258
Copy link

martin0258 commented Apr 10, 2023 •
edited
Loading

@sangdv As far as I know, it is currently not officially supported. However, if you want to implement it by modifying the source code, you may follow ref1 or ref2 as a starting point.

@schliffen
Copy link

Pose (or point) auto annotation is very critic. It would be appreciated if you add this feature too.

@hardikdava
Copy link
Contributor

Hello, I was able to get it working for landmark as type of "points". But I am wondering how to work with type=skeleton with serverless functions. @nmanovic where can I find more information about return type of skeleton for serverless function response?

@hardikdava hardikdava mentioned this issue Jul 17, 2023
Closed
@zhiltsov-max zhiltsov-max mentioned this issue Jul 18, 2023
Merged
5 tasks
@bsekachev bsekachev assigned bsekachev and unassigned sizov-kirill Oct 16, 2023
@bsekachev bsekachev mentioned this issue Oct 19, 2023
Merged
9 tasks
@AnnaPetrovicheva AnnaPetrovicheva mentioned this issue Oct 27, 2023
Closed
nmanovic pushed a commit that referenced this issue Nov 6, 2023
@bsekachev
Added skeleton keypoints detector COCO Whole Body (#7033)
630770e 
### Motivation and context
Resolved #3756
Resolved #5324

Used model is UBody via mmpose
https://mmpose.readthedocs.io/en/latest/model_zoo/wholebody_2d_keypoint.html#topdown-heatmap-hrnet-ubody-coco-wholebody-on-ubody2d

Optional: 
- [ ] Try different detectors from mmdetect
- [ ] GPU support (it very quick on CPU as well)

**Deploy with**: ```./deploy_cpu.sh pytorch/mmpose/ubody2d/nuclio/```

**Recommendations**: redeploy this list of functions if you use any of
them after using this code
 
- tensorflow/faster_rcnn_inception_v2_coco/nuclio/
- tensorflow/matterport/mask_rcnn/nuclio/
- pytorch/facebookresearch/detectron2/retinanet_r101/nuclio/
- openvino/omz/public/yolo-v3-tf/nuclio/
- onnx/WongKinYiu/yolov7/nuclio/
- openvino/omz/public/mask_rcnn_inception_resnet_v2_atrous_coco/nuclio/
-
openvino/omz/public/faster_rcnn_inception_resnet_v2_atrous_coco/nuclio/
- openvino/omz/intel/text-detection-0004/nuclio/
- openvino/omz/intel/semantic-segmentation-adas-0001/nuclio/
- openvino/omz/intel/face-detection-0205/nuclio/
@KTXKIKI
Copy link

KTXKIKI commented Nov 7, 2023

您好，我能够让它作为“点”类型为地标工作。但是我想知道如何使用无服务器函数。在哪里可以找到有关无服务器函数响应的返回类型的详细信息？type=skeleton``skeleton

Now that we support YOLOv8, we can proceed with it

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@bsekachev bsekachev

Labels
models
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Added skeleton keypoints detector COCO Whole Body cvat-ai/cvat
8 participants
@martin0258 @schliffen @sangdv @hardikdava @bsekachev @nmanovic @sizov-kirill @KTXKIKI