Stacking-Based Ensemble Learning Method for the Recognition of the Pedestrian Crossing Intention

Figure 3

Stacking ensemble learning architecture: the data are divided into the training set and the test set. The training set is divided into four training subsets and one validation subset, and a new subset is obtained by the basic classifier RF, SVM, LSTM, and AT Bi LSTM. The new subset is trained by the meta-classifier to obtain the pedestrian crossing intention recognition model. Similarly, the new test set is obtained by four base classifiers. The test set is input into the intention recognition model to obtain the final recognition accuracy.