A continuous sign language recognition and translation dataset with handshape annotations.
It’s an extension of the PHOENIX14T dataset.
PHOENIX14T dataset consists of German sign language (DGS) aired by the German public TV station PHOENIX in the context of weather forecasts. The corpus comprises DGS videos from 9 different signers, glosses annotated by deaf experts, and translations into spoken German language.
We have enriched the SLR dataset PHOENIX14T by incorporating handshape labels derived from the SignWriting dictionary and manual labeling.
Annotations: link
Videos: link
One example from PHOENIX14T-HS is shown below.
When loaded into a Python environment, this example appears as a Python dictionary:
{name: train/01April_2010_Thursday_heute-6703,
signer: Signer04,
gloss: MORGEN TEMPERATUR ACHT BIS DREIZEHN MAXIMAL DREIZEHN,
handshape-right: [[1], [f], [3], [index], [3], [b_thumb], [3]],
handshape-left: [[], [f], [5], [], [], [b_thumb], []]}
Implementations of SLR systems enhanced by handshape annotations from PHOENIX14T-HS can be found in GitHub repo.