LLaSE-G1 A FOSS speech enhancement model

LLaSE-G1 is a unified speech enhancement model capable of handling multiple tasks without extra task prompts, including:<p><pre><code> Noise Suppression (SE) Target Speaker Extraction (TSE) Packet Loss Concealment (PLC) Acoustic Echo Cancellation (AEC) Speech Separation (SS) </code></pre> huggingface link:<p>&gt; <a href="https:&#x2F;&#x2F;huggingface.co&#x2F;ASLP-lab&#x2F;LLaSE-G1" rel="nofollow">https:&#x2F;&#x2F;huggingface.co&#x2F;ASLP-lab&#x2F;LLaSE-G1</a><p>demo page:<p>&gt; <a href="https:&#x2F;&#x2F;submission-papers.github.io&#x2F;LLaSE-G1-demo-page&#x2F;" rel="nofollow">https:&#x2F;&#x2F;submission-papers.github.io&#x2F;LLaSE-G1-demo-page&#x2F;</a><p>arxiv paper:<p>&gt; <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2503.00493" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2503.00493</a>