# Speech Decoding on sEEG signals 61-word Source: [Speech Decoding on sEEG signals 61-word]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:27:10.395Z Evaluation of Speech Decoding performance on the sEEG signals 61-word dataset, measured using metrics like Accuracy and Ste. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Speech Decoding | | Dataset (including any stated split) | sEEG signals 61-word | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy](): Higher is better; unit: PERCENT. - [Ste](): Lower is better; unit: SCALAR. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Du-IN (mae) | 62.7 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | Du-IN (poms) | 59.18 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | Du-IN (vqvae) | 58.24 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | Du-IN | 56.29 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | Neuro-BERT | 49.51 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | EEG-Conformer | 45.82 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | Du-IN (vqvae+vq) | 44.17 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | DeWave | 32.43 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | CNN-BIGRU | 32.04 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | TS-TCC | 24.85 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | Brant | 12.42 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | LaBraM-PopT | 11.78 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | LaBraM | 11.53 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | Brant | 11.16 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | BrainBERT | 7.5 | Not recorded | 68fa5f6f1e209131ce9e9d4d | | BrainBERT | 6.72 | Not recorded | 68fa5f6f1e209131ce9e9d4d | ### Du-IN (mae) Accuracy: 62.7. Result ID: 698bc7b27a967eddb7715087. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":true,"Model Size":"4.38M"}. This record does not assert independent reproduction. ### Du-IN (poms) Accuracy: 59.18. Result ID: 698bc7b27a967eddb7715088. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":true,"Model Size":"5.18M"}. This record does not assert independent reproduction. ### Du-IN (vqvae) Accuracy: 58.24. Result ID: 698bc7b27a967eddb7715086. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":true,"Model Size":"4.38M"}. This record does not assert independent reproduction. ### Du-IN Accuracy: 56.29. Result ID: 698bc7b27a967eddb7715084. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":false,"Multi-subject pre-training (MS)":null,"Model Size":"4.38M"}. This record does not assert independent reproduction. ### Neuro-BERT Accuracy: 49.51. Result ID: 698bc7b27a967eddb771507c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":false,"Multi-subject pre-training (MS)":true,"Model Size":"2.14M"}. This record does not assert independent reproduction. ### EEG-Conformer Accuracy: 45.82. Result ID: 698bc7b27a967eddb771507b. [Reporting paper]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":false,"Multi-subject pre-training (MS)":null,"Model Size":"2.34M"}. This record does not assert independent reproduction. ### Du-IN (vqvae+vq) Accuracy: 44.17. Result ID: 698bc7b27a967eddb7715085. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":false,"Model Size":"4.38M"}. This record does not assert independent reproduction. ### DeWave Accuracy: 32.43. Result ID: 698bc7b27a967eddb771507d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":false,"Multi-subject pre-training (MS)":null,"Model Size":"5.70M"}. This record does not assert independent reproduction. ### CNN-BIGRU Accuracy: 32.04. Result ID: 698bc7b27a967eddb771507a. [Reporting paper]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":false,"Multi-subject pre-training (MS)":null,"Model Size":"0.54M"}. This record does not assert independent reproduction. ### TS-TCC Accuracy: 24.85. Result ID: 698bc7b27a967eddb7715079. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Token Level":"Region","Pre-trained (PT)":false,"Multi-subject pre-training (MS)":null,"Model Size":"0.32M"}. This record does not assert independent reproduction. ### Brant Accuracy: 12.42. Result ID: 698bc7b27a967eddb7715081. [Reporting paper]() Recorded evaluation configuration: {"Token Level":"Channel","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":true,"Model Size":"69.35M"}. This record does not assert independent reproduction. ### LaBraM-PopT Accuracy: 11.78. Result ID: 698bc7b27a967eddb7715083. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Token Level":"Channel","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":true,"Model Size":"6.85M"}. This record does not assert independent reproduction. ### LaBraM Accuracy: 11.53. Result ID: 698bc7b27a967eddb7715082. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Token Level":"Channel","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":false,"Model Size":"6.85M"}. This record does not assert independent reproduction. ### Brant Accuracy: 11.16. Result ID: 698bc7b27a967eddb7715080. [Reporting paper]() Recorded evaluation configuration: {"Token Level":"Channel","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":false,"Model Size":"69.35M"}. This record does not assert independent reproduction. ### BrainBERT Accuracy: 7.5. Result ID: 698bc7b27a967eddb771507f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Token Level":"Channel","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":true,"Model Size":"43.58M"}. This record does not assert independent reproduction. ### BrainBERT Accuracy: 6.72. Result ID: 698bc7b27a967eddb771507e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Token Level":"Channel","Pre-trained (PT)":true,"Multi-subject pre-training (MS)":false,"Model Size":"43.58M"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.