End-to-End Intracortical Speech Decoding from Neural Activity
Quick Take
This study presents an end-to-end Conformer-based neural decoder for intracortical speech decoding, achieving a character error rate (CER) of 23.80% without external language models. The results indicate that effective character-level decoding is feasible, despite challenges from signal degradation and word boundary segmentation errors, particularly relevant for ALS patients.
Key Points
- Achieved a character error rate of 23.80% on validation data.
- Utilized an end-to-end Conformer-based neural decoder.
- No external language models were used during inference.
- Performance variability linked to inter-session signal degradation.
- Dominant errors stem from incorrect word boundary segmentation.
Article Excerpt
From source RSS / original summaryarXiv:2605. 24313v1 Announce Type: new Abstract: Current high-performing intracortical speech neuroprostheses achieve low word error rates but typically rely on external language models during inference, increasing memory, computation, and latency. In this work, we investigate whether meaningful character-level decoding is achievable without such models. We propose an end-to-end Conformer-based neural decoder trained directly on intracortical recordings from a participant with amyotrophic lateral sclerosis (ALS).
Without any external language model, the system achieves a character error rate (CER) of 23. 80\% on held-out validation data. Analysis shows that performance variability is driven by inter-session signal degradation, while dominant errors arise from incorrect word boundary segmentation. These results demonstrate that effective character-level decoding is possible in a fully end-to-end framework, providing a strong neural signal for downstream linguistic processing.
Reader Mode unavailable (could not extract clean content).
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →Time to REFLECT: Can We Trust LLM Judges for Evidence-based Research Agents?
The reliability of LLM judges for evaluating deep research agents is critically assessed using the REFLECT benchmark.