search

open-mmlab / mmpose

OpenMMLab Pose Estimation Toolbox and Benchmark.
https://mmpose.readthedocs.io/en/latest/
Apache License 2.0
5.59k stars 1.22k forks source link

recognition of points #1994

Closed SylvainArd closed 1 year ago

SylvainArd commented 1 year ago

Hello, in this image : https://www.sylvain-ard.fr/temp/feuille.png do you think that mmpose can recognize the differents red points separately, I precise that the number of lobes of the leaf and so the number of points doesn't vary. Thank you Best regards

ly015 commented 1 year ago

I think it's possible to do this task with typical keypoint detection methods in mmpose as long as the lobe number is constant (which means every keypoint is well-defined) and the bounding box of the leaf is available. Of course, it could be a little tricky because the lobes have very similar appearances.

Hi @Tau-J , could you please share your opinions on this question?

Tau-J commented 1 year ago

If all the data is like this image you showed (no collapsing, self-obscuring), then I think it is fine to treat this task as a face 2d keypoint detection task, because the leaves in this case are rigid bodies just like the faces. Therefore, please feel free to play with MMPose to deal with this task and I'm sure the performance will not be bad if you have enough data.

ly015 commented 1 year ago

BTW, if it works on the leaf lobe detection, will you be willing to add a demo in MMPose? This is a novel application of MMPose and many people would be interested.

SylvainArd commented 1 year ago

Thank you, I will test, in fact these are Alchemilla leaves. It is specific leaves. So you want a demo in your github ? Best regards

Le mer. 22 févr. 2023 à 06:07, Yining Li @.***> a écrit :

BTW, if it works on the leaf lobe detection, will you be willing to add a demo in MMPose? This is a novel application of MMPose and many people would be interested.

— Reply to this email directly, view it on GitHub https://github.com/open-mmlab/mmpose/issues/1994#issuecomment-1439446961, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEWZCR4UQGHTAMTZD4DK4N3WYWNJDANCNFSM6AAAAAAVC6WQ7A . You are receiving this because you authored the thread.Message ID: @.***>

SylvainArd commented 1 year ago

In fact after verifying images the number of lobes vary, sometimes 11, sometimes 9 sometimes 7 Sylvain Ard 0549507724 0778380991 @.*** http://sylvain-ard.fr Entreprise individuelle SIRET : 80079243400022 Appt 26 Bât A Résidence Le Patio 83 rue de la Bugellerie 86000 Poitiers

Le mer. 22 févr. 2023 à 10:24, Sylvain Ard @.***> a écrit :

Thank you, I will test, in fact these are Alchemilla leaves. It is specific leaves. So you want a demo in your github ? Best regards

Le mer. 22 févr. 2023 à 06:07, Yining Li @.***> a écrit :

BTW, if it works on the leaf lobe detection, will you be willing to add a demo in MMPose? This is a novel application of MMPose and many people would be interested.

— Reply to this email directly, view it on GitHub https://github.com/open-mmlab/mmpose/issues/1994#issuecomment-1439446961, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEWZCR4UQGHTAMTZD4DK4N3WYWNJDANCNFSM6AAAAAAVC6WQ7A . You are receiving this because you authored the thread.Message ID: @.***>

SylvainArd commented 1 year ago

huh, do you think it can works ? Sylvain Ard 0549507724 0778380991 @.*** http://sylvain-ard.fr Entreprise individuelle SIRET : 80079243400022 Appt 26 Bât A Résidence Le Patio 83 rue de la Bugellerie 86000 Poitiers

Le mer. 22 févr. 2023 à 11:08, Sylvain Ard @.***> a écrit :

In fact after verifying images the number of lobes vary, sometimes 11, sometimes 9 sometimes 7 Sylvain Ard 0549507724 0778380991 @.*** http://sylvain-ard.fr Entreprise individuelle SIRET : 80079243400022 Appt 26 Bât A Résidence Le Patio 83 rue de la Bugellerie 86000 Poitiers

Le mer. 22 févr. 2023 à 10:24, Sylvain Ard @.***> a écrit :

Thank you, I will test, in fact these are Alchemilla leaves. It is specific leaves. So you want a demo in your github ? Best regards

Le mer. 22 févr. 2023 à 06:07, Yining Li @.***> a écrit :

BTW, if it works on the leaf lobe detection, will you be willing to add a demo in MMPose? This is a novel application of MMPose and many people would be interested.

— Reply to this email directly, view it on GitHub https://github.com/open-mmlab/mmpose/issues/1994#issuecomment-1439446961, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEWZCR4UQGHTAMTZD4DK4N3WYWNJDANCNFSM6AAAAAAVC6WQ7A . You are receiving this because you authored the thread.Message ID: @.***>

jin-s13 commented 1 year ago

In fact after verifying images the number of lobes vary, sometimes 11, sometimes 9 sometimes 7 Sylvain Ard 0549507724 0778380991 @. http://sylvain-ard.fr Entreprise individuelle SIRET : 80079243400022 Appt 26 Bât A Résidence Le Patio 83 rue de la Bugellerie 86000 Poitiers Le mer. 22 févr. 2023 à 10:24, Sylvain Ard @.> a écrit : … Thank you, I will test, in fact these are Alchemilla leaves. It is specific leaves. So you want a demo in your github ? Best regards Le mer. 22 févr. 2023 à 06:07, Yining Li @.> a écrit : > BTW, if it works on the leaf lobe detection, will you be willing to add a > demo in MMPose? This is a novel application of MMPose and many people would > be interested. > > — > Reply to this email directly, view it on GitHub > <#1994 (comment)>, > or unsubscribe > https://github.com/notifications/unsubscribe-auth/AEWZCR4UQGHTAMTZD4DK4N3WYWNJDANCNFSM6AAAAAAVC6WQ7A > . > You are receiving this because you authored the thread.Message ID: > @.> >

We need to make the keypoint definition consistent. Missing keypoints are okay, but different definition will not work.

SylvainArd commented 1 year ago

but if we make the points like this : the first lobe after apex, the second lobe after apex, the third etc ? Sylvain Ard 0549507724 0778380991 @.*** http://sylvain-ard.fr Entreprise individuelle SIRET : 80079243400022 Appt 26 Bât A Résidence Le Patio 83 rue de la Bugellerie 86000 Poitiers

Le mer. 1 mars 2023 à 06:38, Jas @.***> a écrit :

In fact after verifying images the number of lobes vary, sometimes 11, sometimes 9 sometimes 7 Sylvain Ard 0549507724 0778380991 @.** http://sylvain-ard.fr Entreprise individuelle SIRET : 80079243400022 Appt 26 Bât A Résidence Le Patio 83 rue de la Bugellerie 86000 Poitiers Le mer. 22 févr. 2023 à 10:24, Sylvain Ard @*.

> a écrit : … <#m-3352493470248932356> Thank you, I will test, in fact these are Alchemilla leaves. It is specific leaves. So you want a demo in your github ? Best regards Le mer. 22 févr. 2023 à 06:07, Yining Li @.> a écrit : > BTW, if it works on the leaf lobe detection, will you be willing to add a > demo in MMPose? This is a novel application of MMPose and many people would > be interested. > > — > Reply to this email directly, view it on GitHub > <#1994 (comment) https://github.com/open-mmlab/mmpose/issues/1994#issuecomment-1439446961>,

or unsubscribe > https://github.com/notifications/unsubscribe-auth/AEWZCR4UQGHTAMTZD4DK4N3WYWNJDANCNFSM6AAAAAAVC6WQ7A . > You are receiving this because you authored the thread.Message ID: > @.***> >

We need to make the keypoint definition consistent. Missing keypoints are okay, but different definition will not work.

— Reply to this email directly, view it on GitHub https://github.com/open-mmlab/mmpose/issues/1994#issuecomment-1449380670, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEWZCR33KSL5CAILNXRUHFLWZ3OG7ANCNFSM6AAAAAAVC6WQ7A . You are receiving this because you authored the thread.Message ID: @.***>