Speech Event Recognition Model based on Deep Learning for People with Dysarthria
Dysarthria is a problem faced by many patients with specific disorders that cause speakers to have unclear articulation.To better understand the speech events expressed by patients with dysarthria,this paper proposes a new deep learning-based speech event recognition model.The model takes speech fragments as input,preserves the original features of the time series using Gramian corner field,extracts local and global features of the sequences using Conformer,and finally uses ResNet as a classification model.The experimental results on EasyCall corpus dataset show that the model proposed in this paper has good recognition effect.
万方数据
维普
