Search papers, labs, and topics across Lattice.
5
0
4
10
Efficient diacritic restoration in Arabic speech can be achieved with a non-autoregressive method that significantly reduces error rates while simplifying the decoding process.
Choosing the right transcription reference can drastically change the perceived performance of ASR systems, especially in atypical speech scenarios.
SER's noble aspirations of voice-activated healthcare are undermined by datasets that bear little resemblance to real-world emotional expression.
Stuttered-speech research is missing the mark: a new study reveals a significant mismatch between current research priorities and the actual needs of people who stutter.
Tokenizing Arabic text morphologically doesn't necessarily improve how well LLMs generate new word forms, suggesting that LLMs may learn morphological patterns independently of explicit tokenization.
