Adaptive Training for Visual Search Expertise

ScreenADAPT® is an adaptive training solution for enhanced visual detection.

Screeners analyze thousands of images daily, therefore, they have to be visual search experts by being both effective and efficient. Examining such high quantities of imagery also heightens the risk of a mishap being overlooked along the way.

ScreenADAPT® develops core visual search skills faster through an individualized training path that addresses skill deficiencies, immediate feedback on progress and errors, dynamically created images of threat and bag images which keeps training fresh and unpredictable.

To find out more about the adaptive training behind ScreenADAPT® DI sat down with one of our Senior Research Associates, Jesse Flint, to find out more.

What is adaptive training?

JF: Adaptive training is training in which the content presented to the trainee is altered in a way that relates to the responses of the trainee.

How is eye movement detected?

JF: ScreenADAPT® uses EyeTech VT3 eye trackers to detect and interpret the eye movements of trainees. The VT3 is an optical tracking system that works by reflecting infrared light from the eye to create bright spots called glints and then tracking the location of the glints using a video camera.

Can you talk about the value of eye tracking?

JF: Typically, adaptive training software is only able to react to the type of errors that a trainee makes. For example, if a trainee misses detecting a specific type of threat, the training can provide more examples of that type of threat in future sessions. Eye tracking allows ScreenADAPT® to assess the root cause of why errors occur, which provides more and better adaptation possibilities. For example, eye tracking allows ScreenADAPT® to determine if a trainee sees a threat and does not recognize what it is (recognition error) or is missing a threat because they are not scanning all areas of an image (scanning error). This allows the system to determine the specific style of training that would best address the root cause of errors.

How are the instructors using ScreenADAPT® for training purposes?

JF: ScreenADAPT® is currently being evaluated by the TSA to determine where it would have the greatest impact on training. The in-progress study is examining the impacts of ScreenADAPT® training for trainees that have exhibited a pattern of poor performance. After the current study, the TSA will evaluate the use of ScreenADAPT® as a tool for new-hire trainees.

What type of deficiencies can ScreenADAPT® detect?

JF: ScreenADAPT® can detect scanning and recognition errors (see above). In addition, it can detect if trainees have a deficiency with specific types of threats, the orientation in which a threat is presented on the screen, high or low cluttered images, indicating that threats are present when there are no threats (false alarms) and if the trainee can detect some but not all of the threats present in a search.

Besides TSA Screeners, what other jobs could benefit from ScreenADAPT®?

JF: Security-based X-ray screening is conducted at a range of military, non-military government (including state and local), and private facilities. ScreenADAPT® could be utilized by the military to screen contents coming onto a base, by Customs and Border Patrol to scan cargo entering the US, by the US Post Office to scan packages in the US mail system, and by Federal Protective as well as many state and local governments to scan packages entering government buildings. In addition, ScreenADAPT® can also be used for a number of non-security-based X-ray screening tasks, such as food quality inspections at canning facilities and medical X-ray image interpretation.

To find out more about ScreenADAPT® check out the product site.

Terence Zagers
Product Engineer