Search papers, labs, and topics across Lattice.
Saarland University
5
0
6
33
Jointly summarizing and translating long-form spoken content could revolutionize how we handle multilingual information processing.
Models trained on AfriSUD reveal a striking syntax gap, highlighting the inadequacy of existing architectures for capturing the complexities of African languages.
No single TTS model excels across all languages, exposing the limitations of current multilingual approaches in low-resource settings.
Code-switched speech can exploit safety weaknesses in LALMs, achieving jailbreak success rates that challenge current safety protocols.
Yor\`ub\'a NLP gets a boost: a new multi-domain NER dataset and language model reveal the limitations of cross-domain transfer and the power of language-specific pretraining.
