PHOENIX14T

[milk120706]

Hello, Eddie,
I noticed you used the same method with the paper "Progressive Transformer for end-to-end sign language production" to get the 3d skeletal joint coordinates, I also followed this method and generated the coordinates data (360 coordinates including face and body). But the quality of the extracted 3D coordinates is not that good. Can you share your pre-processed PHOENIX14T data with me? I would appreciate it if you could share the pre-processed PHOENIX14T data. Here is my email address: [email protected]. Thank you very much.

[thiagomcoutinho]

If its possible, maybe share a link with the files so everyone can used it, please.

@milk120706 , if you get the files could you share them with us? Thanks.

[Eddie-Hwang]

Hello, thank you for your interest in our work.
I would like to share the processed Phoenix dataset, but I do not have permission to publish the modified data. However, you can still ask the author of "Progressive Transformer for end-to-end sign language production" to share the corrected dataset, and he may have the right to release.

Regards,
Eddie

[milk120706]

Hello, Hwang, thank you very much for your reply.
Actually, I sent an email to the author of "Progressive Transformer for end-to-end sign language production" in July but no reply until now. I noticed that you are updating your code, I am very interested in the new evaluation metric Mean Absolute Error of Joint coordinates ( MAJE) you used, I would like to inquire if you can share the code of MAJE. Thank you so much. I hope you have a great day.

[milk120706]

If its possible, maybe share a link with the files so everyone can used it, please.

@milk120706 , if you get the files could you share them with us? Thanks.

I am sorry, I do not get the files.