Sign Language Segmentation With Temporal Convolutional Networks
Katrin Renz, Nicolaj Stache, Samuel Albanie, Gül Varol
-
SPS
IEEE Members: $11.00
Non-members: $15.00Length: 00:07:25
The objective of this work is to determine the location of temporal boundaries between signs in continuous sign language. Our approach marries 3D convolutional neural network representations with iterative temporal segment refinement to resolve ambiguities between boundary cues. We demonstrate the effectiveness of our approach for on the BSLCorpus, Phoenix2014 and BSL-1K datasets, showing considerable improvement over the prior state of the art and the ability to generalise to new signers, languages and domains.
Chairs:
Soohyun Bae