Teaching Audio Models to Reason: A Unified Framework for Source- and Layer-wise Distillation

Documenti analoghi

• HarmoniFuse: A Component-Selective and Prompt-Adaptive Framework for Multi-Task Speech Language Modeling
  di: Si, Yuke, et al.
  Pubblicazione: (2025)
• GenDistiller: Distilling Pre-trained Language Models based on an Autoregressive Generative Model
  di: Gao, Yingying, et al.
  Pubblicazione: (2024)
• PolySpeech: Exploring Unified Multitask Speech Models for Competitiveness with Single-task Models
  di: Yang, Runyan, et al.
  Pubblicazione: (2024)
• Plugin Speech Enhancement: A Universal Speech Enhancement Framework Inspired by Dynamic Neural Network
  di: Chen, Yanan, et al.
  Pubblicazione: (2024)
• MFSN: Multi-perspective Fusion Search Network For Pre-training Knowledge in Speech Emotion Recognition
  di: Sun, Haiyang, et al.
  Pubblicazione: (2023)