• Ocular drift is traditionally considered as noise in the oculomotor system. However, recent psychophysical evidence suggests that our visual system has a certain level of control over it. Here we examined whether the visual system has access to motor knowledge of ocular drift, like saccades and smooth pursuit.

• By precisely manipulating retinal input in a gaze-contingent manner, we created a special task in which retinotopic information doesn't suggest the stimulus's configuration, so that the only way to distinguish stimuli is to rely on motor knowledge of drift direction.

• Our psychophysical data suggest that the visual system has precise motor estimate of ocular drift. Using computational modeling method, we reveal that this motor signal is integrated with retinal input in a probabilistically optimal way to encode spatial information.

Task modulation on head-eye coordination

• Due to technical difficulties, previous studies on fixational eye movements were conducted in head-immobilized conditions. Here, a custom apparatus allows us to precisely measure eye movement in natural head translations.

• With careful geometry transformation, we reconstructed the retinal image motion during natural head-free fixations in real-world tasks.

• Our data show that eyes compensate for the head in such a way to render task-dependent retinal image motion, which is largely irrespective of head movements and saccade sizes.

spatiotemporal content on the retina delivered by saccades transient

• Saccades not only shift our gaze to the next fixational target, but they also introduce abrupt luminance changes on the retina.

• This specific structure of retinal input in space and time interacts with spatiotemporal characteristics of sensory neurons and reshapes visual sensitivity.

• In this study, we applied spectrum analysis to investigate the specific spatiotemporal structure of retinal input delivered by saccade and their potential effects on visual perception.

• Saccades are fast ballistic eye movements that move the fovea to the target of interest, which are our primary way of acquiring novel visual information. However, in the laboratory environments, stimuli are often introduced to subjects by a flash. Is a saccade equivalent to a flash? If not, is the difference led by extra-retinal modulation or how stimuli move on the retinal during saccades?

• Here we measured post-saccadic visual sensitivity across several well-designed experiments to investigate those questions.

• Integrate Scleral search coil with Optitrack, achieving synchronized high-resolution tracking of gaze, head, hand, objects.

• Use machine learning techniques to calibrate and compensate for the inhomogeneity of the magnetic field based on independent measurements from the motion tracking system.

Algorithm for dual-purkinje eye tracker

• Design algorithm for high acuity video-based dual purkinje eye tracker, based on optics of human's eye.