Included are the raw data and the program for classifying each trial to a stimulus target. The classification process includes two stages: a training stage and a testing stage. In the training stage, the raw data derived from a reference template of each of the four target groups are used for creating a spatial filter with canonical correlation analysis (CCA) and 16 templates for the group;In the testing stage, the single-trial data are classified to one of 64 targets using the four spatial filters and the 64 templates yielded in the training stage through template matching. Thus, in the program folder, the following two matlab files need to run for classifying the testing data. 1. run the Matlab file classification_training_four.m to create a spatial filter and 16 templates for each of the four target groups using the training data saved in the data folder; 2. run the Matlab file classification_offlineTest_four.m to classify a single trial to one of the 64 targets using the four spatial filters and 64 templates yielded in the training stage and save in the same folder as matlab codes. The testing data are put in the same folder as the training data.