Why the study?
Existing EEMD methods for EEG emotion recognition generate spurious modes from residual noise and yield varying numbers of decomposed intrinsic mode functions across signals.
Population
19 healthy participants, mean age 22.4 years, participating in a speech-evoked emotion cognitive experiment…
Design
Other
Key result
The proposed DEEMD-SPP framework achieved a classification accuracy of 74.5% for valence and 72.2% for arousal, outperforming other evaluated methods for EEG-based emotion recognition.
Authors
Loading...
May advance EEG emotion recognition tools; leaves open validation for cardiovascular or psychiatric use.
Absolute Event Rate: 74.5% vs 70.8%
The DEEMD-SPP framework improves emotion recognition from EEG signals by effectively eliminating residual noise and selecting valuable intrinsic mode functions.
Chen et al. (2022) studied Emotion recognition (n=19). DEEMD-SPP framework vs. EEMD-SVM was evaluated on Classification accuracy of valence. The proposed DEEMD-SPP framework achieved a classification accuracy of 74.5% for valence and 72.2% for arousal, outperforming other evaluated methods for EEG-based emotion recognition.