Phoneme- vs. Character-Level Targets and Selective State-Space Models for Intracortical Brain-to-Text
This paper compares the performance of recurrent decoders and selective state-space models (Mamba) in intracortical brain-to-text systems, and investigates the impact of output targets (phonetic vs. character) on their performance.
The paper investigates the impact of selective state-space models and output targets on the performance of intracortical brain-to-text systems.
Before reading this…
Applications
- →Assistive technology for individuals with motor disabilities
To understand this paper, make sure you know these concepts first:
- Understanding of neural sequence decodersfind papers →
- Knowledge of state-space modelsfind papers →
Abstract
More Like ThisState-of-the-art intracortical brain-to-text systems pair a neural-sequence phone decoder with an external language model. Two design axes remain underexplored: whether selective state-space models (Mamba) improve on recurrent decoders, and how the output target (phonetic vs.\ character) interacts with that choice. On the public Brain-to-Text '25 benchmark, we study a controlled 2x2 grid (GRU vs.\ hybrid Mamba decoder; phonetic vs.\ character targets) trained with a CTC objective under one reproducible protocol. The recurrent baseline remains strongest: the best phonetic GRU reaches 12.62\% PER and 21.19\% WER, while the best textual GRU after LM rescoring reaches 13.39\% CER and 26.28\% WER. The Mamba hybrid is competitive but does not surpass it. Ablations isolate architectural contributions, and error analysis shows representation-dependent failures: articulatory-like phoneme confusions vs.\ lexical and word-boundary errors.