Similar Face Search for Stronger Security
Similar face search by Kipod is aimed at detecting a person and identifying the events, in which the person was involved. Similar face search is used in security purposes as it allows to detect even unrecognized faces by comparing them to the ones in the database.
Kipod video analytics module includes face recognition, which is widely used for security purposes. Based on deep learning neural networks, the module can identify and recognize faces by comparing them to the ones, stored in the database. As well face recognition module is able to support wide range of turning angles for face identification and the system notifies the user in case a person tries to hide his/her face.
One of face recognition branches is similar face search, based on biometric engine. It allows the operator to perform instant search of a person (either recognized or unrecognized) and see, in which events he/she was involved. Search results also display location and time of the event, when the person was detected, which is a great asset to operators’ work.
One of the greatest advantages of Kipod similar face search is that it can be performed in 2-3 clicks. Upon opening the archive, the user clicks on the person and then, by clicking “search similar faces”, he receives all search results. In the corner of the photo the user will see time and location, when the person was detected.
Kipod technology for similar face search offers few important tools for the user. Firstly the user can define the level of similarity threshold, from 1 to 100% similarity. This feature is highly useful in case the system is used to secure objects with large people masses (airports, i.e.) and the user needs to quickly find a suspect who might endanger the object safety (i.e. terrorist). In this case the user may want to choose 40% threshold, as he/she will need to check every person who even slightly reminds of a suspect.
Another feature is that the user can search for the face not only in the whole database, but according to the lists that he/she created. The lists may be, for example, work departments, so the user can search for the person only within finance or marketing department. This would significantly narrow down the list of faces and provide accurate results.
As already mentioned, the system is able to provide full information about person’s location and time of detection in few clicks. Such efficiency is highly valued among users as it allows them to instantly detect the person and in case of alarm events, take immediate action.
Similar search face by Kipod serves as solution for objects security and person monitoring. Because the system takes into consideration different turning angles of the face, it is able to provide accurate results upon completing the search.
Image taken from http://www.cinemablend.com/television/Iron-Man-Character-Coming-Agent-Carter-67346.html