Rare Word Recognition and Translation Without Fine-Tuning via Task Vector in Speech Models

Documenti analoghi

• High-Fidelity Generative Audio Compression at 0.275kbps
  di: Ma, Hao, et al.
  Pubblicazione: (2026)
• $\text{M}^3\text{PDB}$: A Multimodal, Multi-Label, Multilingual Prompt Database for Speech Generation
  di: Zhu, Boyu, et al.
  Pubblicazione: (2025)
• AudioSpa: Spatializing Sound Events with Text
  di: Feng, Linfeng, et al.
  Pubblicazione: (2025)
• Selective Invocation for Multilingual ASR: A Cost-effective Approach Adapting to Speech Recognition Difficulty
  di: Xue, Hongfei, et al.
  Pubblicazione: (2025)
• Speech Emotion Recognition Using Fine-Tuned DWFormer:A Study on Track 1 of the IERPChallenge 2024
  di: Wang, Honghong, et al.
  Pubblicazione: (2025)