search

orhir / PoseAnything

A Graph-Based Approach for Category-Agnostic Pose Estimation [ECCV 2024]
https://orhir.github.io/pose-anything/
Apache License 2.0
308 stars 17 forks source link

Baseline checkpoints? #8

Closed scfrank closed 8 months ago

scfrank commented 8 months ago

Hi - very cool work, & thanks for releasing the code!

For my own project - CAPE-adjacent, but working with a different dataset and task - I'm more interested in your improved baseline of CapeFormer. Would you be willing to share those training checkpoints - i.e. the base_ config models, rather than the graph_ config models?

Thanks a lot! My university email is st[dot]fr at diku.dk if that's preferable.

Stella

orhir commented 8 months ago

Sure, we can provide the checkpoints. Which model are you interested in - the tiny or small version?

scfrank commented 8 months ago

Wonderful! The small version for now, but I might be interested in the tiny version later.

I think the configs for small_base are already in the repo (I am not looking at the code right now though).

Den søn. 11. feb. 2024 kl. 12.17 skrev Or Hirschorn < @.***>:

Sure, we can provide the configs and checkpoints. Which model are you interested in - the tiny or small version?

— Reply to this email directly, view it on GitHub https://github.com/orhir/PoseAnything/issues/8#issuecomment-1937592150, or unsubscribe https://github.com/notifications/unsubscribe-auth/AADFJLYVMKSXPBHMOR72NMLYTCSE7AVCNFSM6AAAAABDBYRWKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMZXGU4TEMJVGA . You are receiving this because you authored the thread.Message ID: @.***>

orhir commented 8 months ago

Sure. Sent the checkpoints by mail. Here is the link: https://drive.google.com/file/d/1Dj9FE4ZVpo1LbBdYtPvglnQv8uTcz8Ul Closing the issue, if there is anything else needed fell free to open a new one.