yolo's Pose/keypoint detection

[1369355119]

What should I do if I want to use yolo's Pose/keypoint detection? thank you

[surendramaran]

That will be great to implement for mobile devices, sorry but right now my priority is different, I will surely put my time in maintaining this repository as soon as I get some time.

On Thu, 26 Sep, 2024, 8:10 AM 1369355119, @.***> wrote:

Assigned #8 https://github.com/surendramaran/YOLO/issues/8 to @surendramaran https://github.com/surendramaran.

— Reply to this email directly, view it on GitHub https://github.com/surendramaran/YOLO/issues/8#event-14408377916, or unsubscribe https://github.com/notifications/unsubscribe-auth/APXHH4NVFMDC4QQCWJALSULZYNX25AVCNFSM6AAAAABO37OQ3CVHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJUGQYDQMZXG44TCNQ . You are receiving this because you were assigned.Message ID: @.***>

[1369355119]

Hello, I'm glad to get your reply. So far I have developed a version of keypoint/pose on my own. But I'm not sure it's true, especially for the output of the model.

My pose model output is [1,11,8400], which trains only one class, with two key points.

So when I analyze, I think that these 11 channels are as follows: 0 cx 1 cy 2 w 3 h 4 obj_conf 5 kp1_x 6 kp1_y 7 kp1_conf 8 kp2_x 9 kp2_y 10 kp2_conf

But I don't know if that's really true.

So I want to know what to do with the output of the analytic model? Are there any standard methods for accurate analysis?

这对移动设备实施来说会很棒，抱歉，但现在我的 优先级不同，我一定会花时间维护这个 存储库。 … On Thu, 26 Sep, 2024, 8:10 AM 1369355119, @.> wrote: Assigned #8 <#8> to @surendramaran https://github.com/surendramaran. — Reply to this email directly, view it on GitHub <#8 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/APXHH4NVFMDC4QQCWJALSULZYNX25AVCNFSM6AAAAABO37OQ3CVHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJUGQYDQMZXG44TCNQ . You are receiving this because you were assigned.Message ID: @.>