Systems and methods for optimizing operator-state detection including tracking position of an operator-facing camera are described. Systems and methods include receiving a first image captured by an operator-facing camera, detecting a first position of the operator-facing camera with respect to the calibration object, ascertaining the first position with respect to at least one fiducial marker within a passenger compartment of a vehicle, capturing, via controller, a second image using the operator-facing camera, determining the second image is captured by the operator-facing camera from a second position with respect to the at least one fiducial marker, and analyzing, based on determining the second image is captured by the operator-facing camera from the second position, the second image to identify facial features of the operator. The receiving, detecting, ascertaining, determining, and analyzing are performed via a controller. The first image includes a calibration object disposed at a predetermined location.