PHONSSM achieves 72.1% accuracy on a 5565-sign ASL dataset using only skeleton data by enforcing phonological decomposition, outperforming prior skeleton methods by 18.4 points and enabling strong few-shot and zero-shot transfer.
This is expected since components are correlated in natural signs (e.g., certain handshapes occur more often at certain locations)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
State Space Models are Effective Sign Language Learners: Exploiting Phonological Compositionality for Vocabulary-Scale Recognition
PHONSSM achieves 72.1% accuracy on a 5565-sign ASL dataset using only skeleton data by enforcing phonological decomposition, outperforming prior skeleton methods by 18.4 points and enabling strong few-shot and zero-shot transfer.