Eye gaze pattern analysis for fatigue detection based on GP-BCNN with ESM.

•A robust fatigue detection system is constructed by introducing binocular consistency and artificial modulation into CNN.•BCNN introduces binocular consistency constraint into multi-stream network through information interaction module.•GP-BCNN incorporates dual artificial modulation into BCNN to guide the network to learn in a faster and better way.•ESM is proposed to eliminate the detected errors caused by the occluded eyes when the lateral face is detected.•GP-BCNN with ESM obtains the state-of-the-art results and has the generalization potential in general recognition tasks.