-
Notifications
You must be signed in to change notification settings - Fork 32
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.
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
Visibility score request #2
Comments
Hi Sarah, Yes, you are right. We actually used the confidence rate to represent the "probability of each keypoint being visible". You can find that the confidence rate is low when a keypoint is invisible. Tianlang |
Hello and congrats for your work, Like Sarah, I am, actually, trying to generate the visibility score with AlphaPose in order to apply your method with my own videos. George Andrei |
Greetings, I am also wondering about this as well. We are a group of researchers, trying to implement and train this great paper on our own videos. I would be grateful if you could provide some information about the score.pkl, the generation and the format you have used in your work. Thank you in advance. Sincerely, |
Hi,
The score.pkl contains the 17 2D-keypoints’ visibility confidence of each H3.6M video frame predicted by AlphaPose (https://github.com/MVIG-SJTU/AlphaPose). It should be noticed that these confidence-related keypoints predicted by AlphaPose are in the COCO-format which is slightly different from the 2D keypoint input extracted by Cascaded Pyramid Network (e.g. the COCO-format contains keypoints of eyes/ears but does not contain keypoint of neck). However, it still improves the performance. We use the networks described in Section III-E to utilize the keypoints’ visibility confidence information.
Best wishes,
Tianlang
From: Amin Ansarian <notifications@github.com>
Sent: 2021年2月16日 11:31
To: sunnychencool/Anatomy3D <Anatomy3D@noreply.github.com>
Cc: Chen, Tianlang <tchen45@cs.rochester.edu>; Comment <comment@noreply.github.com>
Subject: [EXT] Re: [sunnychencool/Anatomy3D] Visibility score request (#2)
Greetings,
I am also wondering about this as well. We are a group of researchers, trying to implement and train this great paper on our own videos. I would be grateful if you could provide some information about the score.pkl, the generation and the format you have used in your work. Thank you in advance.
Sincerely,
Amin Ansarian
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_sunnychencool_Anatomy3D_issues_2-23issuecomment-2D779954008&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=rPOBJC5gKaoaPCPUTVTptPi9X8fSUVufMYyYR7i8ztA&m=9nAqfpT0hplWoPjPT2d79bGkU1X4BPY22UbSZQmdmMI&s=bOUxi3duxXFU2DkKzZywwlCrNInCNDtjx-jbAKfkPsY&e=>, or unsubscribe<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AFQKPEFTZAWBUNUFNQ6G7P3S7KMURANCNFSM4NE5IYKQ&d=DwMCaQ&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=rPOBJC5gKaoaPCPUTVTptPi9X8fSUVufMYyYR7i8ztA&m=9nAqfpT0hplWoPjPT2d79bGkU1X4BPY22UbSZQmdmMI&s=_nkW1F2C7F4PZNjhwQWDxU6MJGbSx7lAgQklnzOjXUs&e=>.
|
Thanks for your great contributions on this paper. I also want to know more details about the score.pkl mentioned in in Section III-E which utilize the keypoints’ visibility confidence information using AlphaPose. |
Dear Dr. Tianlang,
I am, actually, trying to generate the visibility score with Alphapose in order to apply your method with my own videos.
I will be deeply thankful if you can give me some details about how you generate the file score.pkl ?
In fact, concerning the visibility score, in your paper it’s defined as “[…] the probability of each keypoint being visible in the frame […] ”. I am somehow confused about this issue. Could we say that it’s the confidence rate used in Alphapose ?
I want to sincerely thank you for the response and for your kind help,
Sarah Solaiman
The text was updated successfully, but these errors were encountered: