Search papers, labs, and topics across Lattice.
The authors address stress-aware sentence-level grapheme-to-phoneme (G2P) conversion for Filipino by fine-tuning a multilingual ByT5 model on weakly-supervised corpora generated via an LLM-assisted pipeline guided by Wiktionary. Sentence context is essential to disambiguate prosodic stress and lexical homographs in Filipino, yet sentence-level phonemic annotations are virtually non-existent in available datasets. On a manually validated test set, their approach slashes phoneme error rate from 19.74% to 0.54% (and character error rate to 2.50%), successfully capturing major stress classes while highlighting residual weaknesses on glottal *malumi* forms.
Weakly-supervised LLM annotations can bridge the phonemic data gap in low-resource languages, slashing Filipino sentence-level G2P error rates from nearly 20% down to 0.54% while successfully disambiguating prosodic homographs.
Grapheme-to-phoneme conversion (G2P) refers to the task of converting a sequence of graphemes to a corresponding sequence of phonemes. While Filipino G2P is fairly straightforward due to its shallow orthography, the inclusion of prosodic features such as stress adds a layer of complexity that requires sentence-level context instead of single-word inputs. However, sentence-level data for Filipino typically do not include phoneme transcriptions, posing a challenge for training G2P models. As such, we investigate how to obtain sentence-level phoneme data for Filipino using available data and compare the resulting models with multilingual word-level G2P as well as measure how accurately they predict stress marker position for Filipino. We propose fine-tuning a ByT5-based model, pre-trained on multilingual word-level G2P data, on three sentence-level G2P datasets annotated with an LLM-assisted pipeline guided by data from Wiktionary. This approach produces models that perform well on the G2P task, achieving at best around 0.54% PER and 2.50% CER, a significant decrease compared to base model PER at around 19.74%, on a manually-corrected test set. The model is able to correctly classify most of the main stress classes in Filipino, but struggles particularly with malumi words. We show that a ByT5-based model performs well at sentence-level Filipino G2P and offers strong potential for Filipino homograph disambiguation.
