Quantifying Human Performance of a Dynamic Military Target Detection Task: An Application of the Theory of Signal Detection.
Abstract
As crew aiding technologies are developed to assist military aviators in performing complex target detection tasks, evaluation metrics must be developed which are common to both human operators and automatic target recognition (ATR) systems so that performance comparisons can be efficiently conducted. The dynamic nature of the multiple target detection task introduces several unique problems in quantifying detection performance. Classical methods of implementing the Theory of Signal Detection (TSD) to quantify performance have proven to be insufficient, and ATR evaluators have developed unique metrics which have not been applicable to evaluating human performance. This research introduced a novel application of TSD to the dynamic, multiple target detection scenario, and a new method of evaluating human performance was developed by adapting an established ATR evaluation method to human subject performance. A linear relationship was discovered between the TSD metric d' and the new evaluation metric, validating the new evaluation method. The new method provided a common metric for evaluating both human and ATR performance of multiple target detection tasks.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1995
- Accession Number
- ADA308392
Entities
People
- Stuart L. Turner
Organizations
- Armstrong Laboratory