This repository contains the HandArch architecture for sign language hand pose recognition.
This architecture was presented and publish in the proceedings of Workshop on computer vision (WVC) 2021 with the title:
"HandArch: A deep learning architecture for LIBRAS hand configuration recognition"
⭐ Conference BEST paper award.
The presentation can be viewed in WVC's Youtube channel
⚠️ This repository is not yet functional, it contains the bare minimum code to create a sign language recognition system and lacks some methods. We are working to finish this repository with installation/running requirements and sample code to demosntrate usage.
We are still working in a way to make the datasets available. IF you want the datasets, please, send me an e-mail with title "LIBRAS91 Dataset request" and I will send a link to download it.