UniAE-MoE: A Unified Audio Encoder via Mixture of Experts

arXiv:2609.39199 · cs.SD, cs.AI · Submitted 2026-09-30 · Read on arXiv

cs.SD, cs.AI

Submitted: 2026-09-30

Updated: 2026-10-05

Code: https://github.com/Syclus123/UniAE-MoE

Terminology

• Mixture of Experts — 1× in this paper · explained in 2 episodes

Sources

• MusicLM: Generating Music From Text
• SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model
• Fun-ASR Technical Report
• Qwen2-Audio Technical Report
• FMA: A Dataset For Music Analysis
• Kimi-Audio Technical Report
• Audio Flamingo 3: Advancing Audio Intelligence with Fully Open Large Audio Language Models
• MoE Adapter for Large Audio Language Models: Sparsity, Disentanglement, and Gradient-Conflict-Free
• MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Training
• AudioPaLM: A Large Language Model That Can Speak and Listen
• GLU Variants Improve Transformer
• Qwen3-Omni Technical Report
• Qwen2.5-Omni Technical Report

Related papers

• Few-Shot Open-Set Audio Classification via Transductive Prototype Refinement and Class Logit Enhancement
• Spectral Masking and Interpolation Attack (SMIA): A Black-box Adversarial Attack against Voice Authentication and Anti-Spoofing Systems
• Remix the Timbre: Diffusion-Based Style Transfer Across Polyphonic Stems
• AVMeme Exam: A Multimodal Multilingual Multicultural Benchmark for LLMs' Contextual and Cultural Knowledge and Thinking
• SoundWeaver: Compositional Warm-Starting for Text-to-Audio Diffusion Serving
• WASIL: In-the-Wild Arabic Spoken Interactions with LLMs