SEAL: Mixture-Closed Additive Reconstruction and Refinement-Aware Expert Routing for Efficient Speech Separation
On the map: Mixture-of-experts → Multimodal MoE models also fits: Expert routing strategies
Abstract (arXiv)
Compact time-frequency separators that mask the mixture and refine through a shared cell face two limits. First, a bounded multiplicative mask only scales a mixture bin, so where overlapping components cancel, the estimate stays small. Second, a shared cell applies the same weights to every time-frequency token at every step, so enlarging it adds compute everywhere. We present SEAL (Sparse Expert routing with Additive Latent reconstruction) to address both. For reconstruction, a zero-sum additive residual bounded by the local mixture amplitude lets estimates be nonzero where components cancel yet still sum to the mixture. For routing, a query built from acoustic and inter-step evidence sends each token to one of six residual experts, and a norm cap keeps the step cue from overriding clear acoustic evidence. On EchoSet, SEAL (small) surpasses TIGER (small) by 0.31 dB SI-SDRi with 28% fewer parameters and 2.9 times fewer MACs, and SEAL (large) is within 0.07 dB SI-SDRi of TIGER (large) at 3.1 times fewer MACs.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Multimodal MoE models, one of 12 directions in Mixture-of-experts. Builds on 2 of its 30 references that are in the map. First posted on arXiv 5 Oct 2026.
Builds on
2 of 30 references.
Closest ancestors · 1
Same direction, newest first
Other references · 1
- 2022ST-MoE: Designing Stable and Transferable Sparse Expert Modelscited by 399
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026TF-MoE: Time-Frequency Mixture-of-Experts for Efficient Speech Separation
- 2026Stride-k Subsampling: Train-Free Audio Token Reduction for Whisper
- 2025Omni-Router: Sharing Routing Decisions in Sparse Mixture-of-Experts for Speech Recognition
- 2022Parameter-Efficient Conformers via Sharing Sparsely-Gated Experts for End-to-End Speech Recognition
- 2024Mixture Compressor for Mixture-of-Experts LLMs Gains More
- 2026Enhancing Flow Matching with A Unified Guidance Framework for Efficient and Robust Speech Synthesis
- 2026Shared Low-rank Basis Factorization for Data-free Mixture-of-Experts Compression
- 2026OMP-MoE: Efficient Expert Pruning for Mixture-of-Experts LLMs via Orthogonal Matching Pursuit