Improving Language Identification for Multilingual Speakers
This addresses a critical gap for multilingual users in language identification systems, though it is incremental as it builds on existing methods with context integration.
The paper tackled the problem of language identification for multilingual speakers, where existing systems underperform on accented speech, by combining a coarse-grained acoustic model with interaction context signals, achieving 97% average accuracy and over 60% relative improvement in worst-case accuracy.
Spoken language identification (LID) technologies have improved in recent years from discriminating largely distinct languages to discriminating highly similar languages or even dialects of the same language. One aspect that has been mostly neglected, however, is discrimination of languages for multilingual speakers, despite being a primary target audience of many systems that utilize LID technologies. As we show in this work, LID systems can have a high average accuracy for most combinations of languages while greatly underperforming for others when accented speech is present. We address this by using coarser-grained targets for the acoustic LID model and integrating its outputs with interaction context signals in a context-aware model to tailor the system to each user. This combined system achieves an average 97% accuracy across all language combinations while improving worst-case accuracy by over 60% relative to our baseline.