Open neouyghur opened 1 year ago
Hi, VLPart paper is under company internal review and it will be released in about one month. Please stay tuned~
Hi,
Does it support part segmentation without object category such as 'head' (not dog head
)? another question what is the difference between Grounded-Segment-Anything
and this project? thanks.
Hi, For now, it doesn't support the general part concept like 'head', since 'dog head' and 'person head' are largely different.
Grounded-Segment-Anything is Grounded DINO + SAM, and this project is GLIP + SAM. Actually, any open-vocabulary (text-guided) object detection model can be used to combine with SAM.
From: Peize Sun @.> Sent: Wednesday, April 12, 2023 4:52:47 PM To: Cheems-Seminar/segment-anything-and-name-it < @.> Cc: Osman Tursun @.>; Author @.
Subject: Re: [Cheems-Seminar/segment-anything-and-name-it] About VLpart (Issue #1)
Hi, For now, it doesn't support the general part concept like 'head', since 'dog head' and 'human head' are largely different.
Grounded-Segment-Anything is Grounded DINO + SAM, and this project is GLIP
— Reply to this email directly, view it on GitHub https://github.com/Cheems-Seminar/segment-anything-and-name-it/issues/1#issuecomment-1504756874, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABQRQ3CQBJ457L5ETWGX5UTXAZGL7ANCNFSM6AAAAAAW2BWUBQ . You are receiving this because you authored the thread.Message ID: @.***>
@PeizeSun 请问VLPart支持分割人类手指等细粒度的肢体吗 我在测试了很多张实验都没有成功分割出来。使用的prompt有什么需要注意的吗
@PeizeSun 请问VLPart支持分割人类手指等细粒度的肢体吗 我在测试了很多张实验都没有成功分割出来。使用的prompt有什么需要注意的吗
VLPart does not support part segmentation on human fingers due to lack of such data in training.
Thank you for your patient reply.
Hi, I am interested VLpart model. Where can I find the paper? Thanks.