Phonological Representation Learning for Isolated Signs Improves Out-of-Vocabulary Generalization
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918136097275904 |
|---|---|
| author | Kezar, Lee Sehyr, Zed Thomason, Jesse |
| author_facet | Kezar, Lee Sehyr, Zed Thomason, Jesse |
| contents | Sign language datasets are often not representative in terms of vocabulary, underscoring the need for models that generalize to unseen signs. Vector quantization is a promising approach for learning discrete, token-like representations, but it has not been evaluated whether the learned units capture spurious correlations that hinder out-of-vocabulary performance. This work investigates two phonological inductive biases: Parameter Disentanglement, an architectural bias, and Phonological Semi-Supervision, a regularization technique, to improve isolated sign recognition of known signs and reconstruction quality of unseen signs with a vector-quantized autoencoder. The primary finding is that the learned representations from the proposed model are more effective for one-shot reconstruction of unseen signs and more discriminative for sign identification compared to a controlled baseline. This work provides a quantitative analysis of how explicit, linguistically-motivated biases can improve the generalization of learned representations of sign language. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_04745 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Phonological Representation Learning for Isolated Signs Improves Out-of-Vocabulary Generalization Kezar, Lee Sehyr, Zed Thomason, Jesse Computation and Language Computer Vision and Pattern Recognition Sign language datasets are often not representative in terms of vocabulary, underscoring the need for models that generalize to unseen signs. Vector quantization is a promising approach for learning discrete, token-like representations, but it has not been evaluated whether the learned units capture spurious correlations that hinder out-of-vocabulary performance. This work investigates two phonological inductive biases: Parameter Disentanglement, an architectural bias, and Phonological Semi-Supervision, a regularization technique, to improve isolated sign recognition of known signs and reconstruction quality of unseen signs with a vector-quantized autoencoder. The primary finding is that the learned representations from the proposed model are more effective for one-shot reconstruction of unseen signs and more discriminative for sign identification compared to a controlled baseline. This work provides a quantitative analysis of how explicit, linguistically-motivated biases can improve the generalization of learned representations of sign language. |
| title | Phonological Representation Learning for Isolated Signs Improves Out-of-Vocabulary Generalization |
| topic | Computation and Language Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.04745 |