In this paper, we propose a cross-subject multi-modal emotion recognition framework, which jointly learns subject-independent representation and common feature between EEG and eye movements. Firstly, ...