List your company
ResearchFormerly WizWand

Speech Decoding on sEEG signals 61-word

62.7AccuracyDu-IN (mae)
4.480819.595434.7149.8246May 19, 2024
Updated 8mo ago

Evaluation results

MethodLinks
Du-IN (mae)Token Level=Region, Pre-trained (PT)=true, Multi-subject pre-training (MS)=true, Model Size=4.38M2024.0562.74.69
Du-IN (poms)Token Level=Region, Pre-trained (PT)=true, Multi-subject pre-training (MS)=true, Model Size=5.18M2024.0559.184.63
Du-IN (vqvae)Token Level=Region, Pre-trained (PT)=true, Multi-subject pre-training (MS)=true, Model Size=4.38M2024.0558.244.83
Du-INToken Level=Region, Pre-trained (PT)=false, Model Size=4.38M2024.0556.295.2
Neuro-BERTToken Level=Region, Pre-trained (PT)=false, Multi-subject pre-training (MS)=true, Model Size=2.14M2024.0549.514.43
EEG-ConformerToken Level=Region, Pre-trained (PT)=false, Model Size=2.34M2024.0545.824.66
Du-IN (vqvae+vq)Token Level=Region, Pre-trained (PT)=true, Multi-subject pre-training (MS)=false, Model Size=4.38M2024.0544.174.04
DeWaveToken Level=Region, Pre-trained (PT)=false, Model Size=5.70M2024.0532.434.48
CNN-BIGRUToken Level=Region, Pre-trained (PT)=false, Model Size=0.54M2024.0532.045.45
TS-TCCToken Level=Region, Pre-trained (PT)=false, Model Size=0.32M2024.0524.854.42
BrantToken Level=Channel, Pre-trained (PT)=true, Multi-subject pre-training (MS)=true, Model Size=69.35M2024.0512.424.1
LaBraM-PopTToken Level=Channel, Pre-trained (PT)=true, Multi-subject pre-training (MS)=true, Model Size=6.85M2024.0511.782.7
LaBraMToken Level=Channel, Pre-trained (PT)=true, Multi-subject pre-training (MS)=false, Model Size=6.85M2024.0511.532.63
BrantToken Level=Channel, Pre-trained (PT)=true, Multi-subject pre-training (MS)=false, Model Size=69.35M2024.0511.163.56
BrainBERTToken Level=Channel, Pre-trained (PT)=true, Multi-subject pre-training (MS)=true, Model Size=43.58M2024.057.51.76
BrainBERTToken Level=Channel, Pre-trained (PT)=true, Multi-subject pre-training (MS)=false, Model Size=43.58M2024.056.721.59