search

GXYM / TextBPN-Plus-Plus

Arbitrary Shape Text Detection via Boundary Transformer；The paper at: https://arxiv.org/abs/2205.05320, which has been accepted by IEEE Transactions on Multimedia (T-MM 2023).
172 stars 37 forks source link

Pretrained weights on Huggingface? #21

Open jturner116 opened 1 year ago

jturner116 commented 1 year ago

Hello authors, Thank you for your great work! I am very interested in your models since they do not rely on OpenMM like many other popular OCR implementations. Would it be possible to host your weights on Huggingface instead of Google Drive? I believe it would allow more people to benefit from your great work.

Thank you! Jonah

GXYM commented 1 year ago

Hello authors, Thank you for your great work! I am very interested in your models since they do not rely on OpenMM like many other popular OCR implementations. Would it be possible to host your weights on Huggingface instead of Google Drive? I believe it would allow more people to benefit from your great work.

Thank you! Jonah

This is a great idea. We will carefully consider your suggestion. But considering that we have no experience using Huggingface. We will first investigate the cost of this process. If feasible, we will place the model on Huggingface. Of course, you can also download our model and code to deploy them on Huggingface, and we will do our best to provide you with necessary authorization and assistance. If you have done so, please let us know the entrance as we would like to experience it. Thank you!

jturner116 commented 1 year ago

Excellent! I believe the cost is free, here is a link describing the model uploading process https://huggingface.co/docs/hub/models-uploading I have not had a chance to upload the model myself, but I might get a chance this weekend if you haven't beat me to it. Thanks! Jonah

GXYM commented 1 year ago

Excellent! I believe the cost is free, here is a link describing the model uploading process https://huggingface.co/docs/hub/models-uploading I have not had a chance to upload the model myself, but I might get a chance this weekend if you haven't beat me to it. Thanks! Jonah

ok, if it works, give us a test link. Thanks.