Gaze determination machine learning system having adaptive weighting of inputs

A method comprising:

• determining, using one or more machine learning models and based at least on image data representative of an image:
• a set of eye direction features associated with a subject;
• a gaze direction associated with the subject;
• and
• face orientation information associated with the subject;
• determining, based at least on the face orientation information and one or more eye direction features of the set of eye direction features, a set of modified eye direction features;
• determining an updated gaze direction associated with the subject based at least on the set of modified eye direction features and the gaze direction;
• and
• initiating an operation based at least on the updated gaze direction.

A system comprising:

• one or more processing units to:
• determine, using one or more machine learning models and based at least on at least a portion of image data representative of an image depicting at least one eye of a subject, a set of features;
• determine, using the one or more machine learning models and based at least on the image data, one or more face orientation features corresponding to an orientation of a face of the subject and a gaze direction associated with the subject;
• determine, based at least on the set of features and the one or more face orientation features, a modified set of features;
• determine, based at least on the gaze direction and the modified set of features, an updated gaze direction associated with the subject;
• and
• initiating an operation based at least on the updated gaze direction.