脑电信号的稳定扩散样本增强方法OA

In response to the classification performance bottleneck caused by the scarcity of EEG signal samples,this paper proposes a stable diffusion model sample enhancement method for EEG.By converting EEG samples into time-frequency maps and using effective cue words to fine tune the image stabilization diffusion model,the EEG sample set is expanded to improve classification and recognition accuracy.The experimental results show that the classification accuracy of the sample set enhanced by the stable diffusion model has been improved from 72.94%to 76.80%,indicating that the proposed method effectively overcomes the classification performance bottleneck and provides a new approach for small sample EEG sample analysis.