Unlocking Animal Pose Estimation with YOLOv8: Fine-tuning for Dogs

[@LearnOpenCV]

Try out our FREE Courses at opencv.org/university/free-courses/

[@Maddy_akil]

Works super good thank you sir

[@RokiFauzi]

can you help me ? error FileNotFoundError: [Errno 2] No such file or directory: 'StanfordExtra_V12/StanfordExtra_v12.json'

[@LearnOpenCV]

Fill this google form: forms.gle/sRtbicgxsWvRtRmUA

[@ingridakaz9097]

Is it possible to estimate x,y and z keypoint coordinates? As I understand YOLOv8 is only capable estimating X and Y?

[@LearnOpenCV]

Right now there are no provisions to estimate 3d coordinates (x, y, and z). One can fine-tune the pose models only with 2D coordinates with an optinaly visibility flag. Please check the documentation.docs.ultralytics.com/datasets/pose/

[@tershire]

Thank you for the video! If the animal is not dog, should I make dataset by myself?

[@LearnOpenCV]

You will either have to search for a dataset, or create one yourself. Check out our playlist on CVAT to know how to create the datasets: ruclips.net/p/PLfYPZalDvZDLvFhjuflhrxk_lLplXUqqB

[@tershire]

@@LearnOpenCV Wonderful. Thank you !

[@RokiFauzi]

If I want to upload an image and produce an output image that already has keypoint annotations using this code? how can i do that ? please answer :)

[@LearnOpenCV]

Please go through the prepare_predictions code for image inference explained in the blog post.

[@RokiFauzi]

@@LearnOpenCV can you write the full code that i mean pleeeaaseee ? i need to upload image and want to show output image that already has keypoint. I need your help

[@LearnOpenCV]

Please check this article for the code: learnopencv.com/animal-pose-estimation/@@RokiFauzi

[@Snower19764]

Where is the Google Colab link? I cannot find it.

[@LearnOpenCV]

Download the code from this blog post: learnopencv.com/animal-pose-estimation/