This research will examine training with an augmented reality learning system to identify combat vehicles. Due to the increase in use of unmanned vehicles (UVs) for missions, a question arises: How do we best train operators to perform well when presented with a combat identification task. More specifically: (a) Is training using canonical (front and side) views sufficient? (b) Due to UAV perspective surveillance, are non-canonical/birds eye views necessary for optimal combat identification performance? (c) Would training with either perspective yield sufficient performance?