Skip to content

MixToken

Home / Concepts / MixToken

MixToken

Auto-generated stub. Edit this file to add more details.

A task-specific mixture-of-experts architecture that fuses signal-specific filterbank features with cross-channel statistical tokens via learned logit mixing.

Why It Matters

It provides a high-performance, parameter-efficient alternative to fine-tuning large foundation models in resource-constrained wearable edge environments.

Evidence

we develop two novel methodologies for hand-gesture recognition: (i) MixToken, a task-specific mixture-of-experts that fuses per-channel IMU filterbank features with cross-channel statistical tokens through learned logit mixing

Metadata & Links