An image capturing apparatus comprising an object detection unit which detects a specific object from an image signal, and a control unit which performs first control corresponding to the specific object when the object detection unit detects the specific object, and performs second control different from the first control when the object detection unit does not detect the specific object, wherein when a state in which the specific object is detected by the object detection unit transits to a state in which the specific object becomes undetectable, the control unit changes, based on information before the specific object becomes undetectable, at least either of a time for which the first control is held and a transition speed when transiting from the first control to the second control.