← Back to Model Beat
Models·2d ago·all news from November 20, 2026

Language-Oriented Semantic Modeling for Target Language Extraction

arXiv:2611.26317v1 Announce Type: new Abstract: Target Language Extraction (TLE) aims to isolate speech in a target language from multilingual mixtures. Existing approaches primarily rely on acoustic modeling with language identity labels, which provide only coarse categorical guidance. In this work, we propose a language-oriented semantic modeling framework that augments acoustic extraction with language-specific semantic representations. A pre-trained speech semantic model is adapted via language-aware prompt conditioning and parameter-efficient fine-tuning to generate target-language-focused semantic representations from the mixture. {These capture high-level semantic structure of the acoustics and they} are integrated into the extraction model to provide structured semantic guidance beyond identity labels. Experiments on the CommonVoiceMix dataset show that incorporating semantic representations consistently improves over acoustic-only baselines, achieving up to 1.16 dB SI-SDR gain.

ModelsExtract

Covered by 1 source

  • AarXiv CS.AI↗Mehmet Sinan Y{\i}ld{\i}r{\i}m, Ke Zhang, Shuai Wang, Haizhou Li2d ago

Related stories

ModelsChina Dismisses Claim that It Illicitly Extracts Foreign AI TechJul 17 · 94 sourcesHardwareSakana AI Expands Focus on Defense and Intelligence Analytics With Narrative-Extraction Technology - TipRanksMar 22PolicyOpenAI makes it much harder for attackers to extract how its latest AI models thinkNov 20ModelsStructured Extraction of Diffraction Data from Covalent Organic FrameworksNov 19