Eye AF

Eye AF uses computational subject recognition to identify eyes within the camera's field of view and prioritise them as the autofocus target. Rather than relying on the photographer to position a focus point over the subject's eye manually, Eye AF scans the scene, detects faces, locates the eyes within detected faces, and places the AF point precisely on the eye. When the subject moves, the camera continues tracking the eye across the frame without requiring any input from the photographer.

In M43, Eye AF is implemented differently across brands. OM System calls their implementation AI Subject Detection, which covers human eyes as part of a broader recognition system that also identifies animals, birds, motorsport vehicles, trains, and aircraft. Panasonic implements human eye detection on the G9 II and GH7 via their Subject Recognition AF, with animal and bird detection also available. Earlier Panasonic bodies had limited or no eye detection, which was a significant disadvantage for portrait photography.

The accuracy and reliability of Eye AF depends heavily on the camera's processing capability and the conditions. Faces partially obscured, faces at extreme angles, low light, and heavy compression artefacts in the scene all reduce detection accuracy. OM System's stacked sensor processing on the OM-1 Mark II maintains eye tracking through brief obstructions — when a subject passes behind something briefly, the camera maintains the track and reacquires the eye when it reappears. For portrait, wedding, and event photography, reliable Eye AF significantly reduces the cognitive load of manual focus point placement.