Multimodal Approach for Emotion Recognition Based on Simulated Flight Experiments

Figure 1. Airplane Cessna 172SP (left); flight route (red line) of the experiment (right).

Figure 2. Face recording of some users during experiment.

Figure 3. Electrodes placement. EEG and HR, placed on the scalp and earlobe (left); and GSR, placed on the indicator and middle fingers (right).

Figure 4. Datasets used in this work.

Figure 5. Poincaré plot demonstration over the flight dataset RC2.

Figure 6. RTOR operation diagram.

Figure 7. Realtime outlier correction based on RTOR method. Note the corrected output (blue) and the raw output with outliers (green).

Figure 8. Errors results (RMSE+MAE) from tests 3–6 (with feature extraction).

Figure 9. Errors results (RMSE) comparison from tests 3–34 (with feature extraction).

Figure 10. Errors results (MAE) comparison from tests 3–34 (with feature extraction).

Figure 11. Errors results comparison between RMSE and MAE from tests 1 to 34 (with feature extraction).

Figure 12. Major emotion accuracies from the tests 3 to 6 (with feature extraction).