search

biaofuxmu / ConSLT

A Token-level Contrastive Framework for Sign Language Translation
MIT License
4 stars 1 forks source link

csl-daily dataset training #1

Open jekim5418 opened 1 year ago

jekim5418 commented 1 year ago

Thank you for your research.

I found that there exist experiments of PHOENIX-2014T and CSL-daily and the model architecture is based on Sign2(Gloss+Text). The sign2 (Gloss+Text) model uses features generated by pretrained spatial embeddings, but there is no feature pickle in the csl-daily dataset. It would be appreciated if you could tell us how the features created by csl-daily were obtained.

biaofuxmu commented 12 months ago

Thank you for your research.

I found that there exist experiments of PHOENIX-2014T and CSL-daily and the model architecture is based on Sign2(Gloss+Text). The sign2 (Gloss+Text) model uses features generated by pretrained spatial embeddings, but there is no feature pickle in the csl-daily dataset. It would be appreciated if you could tell us how the features created by csl-daily were obtained.

We use the features from the file Extracted_Features_with_BN-TIN-Transformer.zip in the CSL-Daily dataset.