Security Industry into the Era of Artificial Intelligence
November 09, 2016
Artificial intelligence should be gone through data, technology, products, the three cycles to complete the process loop. To Google "search engine" as an example, the product is a web page search engine, technology is the algorithm to achieve page rankings, data is generated by the user clicks the page rankings. In the field of security, to open science and technology "deep learning engine", for example, if the product is intelligent visual cloud platform, technology is the depth of learning algorithm, the data is marked image data. Then, artificial intelligence platform is a data and technology through the polished out of the product, the product application process has generated new data, iterative and deepening the learning process. That is, products, technologies and data to form a super-artificial intelligence technology closed-loop, the ultimate goal is to form a machine vision functional effects, giving the machine a pair of eye, so that it like people to "see" and "think." So that the machine has the same visual function as a person, in order to achieve a variety of detection, judgment, recognition, measurement and other functions.
1. Commercial applications of face recognition
As the representative technology of artificial intelligence, the ability of product application and commercialization is very important. In the aspect of accuracy and security, the false recognition rate of less than one hundred thousandth of the accuracy rate is achieved by using technology of face recognition technology as a reference, and the passing rate is over 95%. In the security level, higher than the security level of fingerprint recognition, and no special hardware support needs, more easily be promoted.
In the field of technical application of face recognition, it has realized the concrete business function in the commercial building, and helped the merchant to realize the accurate analysis of the passenger flow from "statistics" to "attribute" to "behavior", to obtain more by machine vision analysis. Multi-line data dimension, with large data analysis to guide the precise marketing business, and promote the retail industry from the same products and services to all people, toward the same product, different services to a specific person to change.
At the same time, the use of face recognition technology to enterprises, parks bring more intelligent management experience. Such as through the face recognition technology products, in the face recognition gates, access control, conference attendance and other traditional scenes bring management efficiency and convenience.
2. Application of Face Recognition in Security Industry
Traditional security industry to video surveillance-based, in the massive unstructured video data, the application of structured data have a lot of opportunities. The Kuang technology as a result of artificial intelligence technology to the traditional security industry, let the machine to understand the world rather than the chaotic video stream, so that the traditional industry to activate more value.
The face recognition technology has gone through the stages of laboratory algorithm, software SDK stage, single product stage, solution stage, and integrated platform stage, based on the time axis analysis of face recognition development. At present, many companies are in the software SDK stage and single product stage, this stage is actually a DEMO stage, open science and technology in the single product stage, the solution stage and the integrated platform stage, the current application areas pan-financial, pan-commercial and pan-security. Pan-security applications in which there are two applications during the G20 summit in Hangzhou, one in the hotel front desk, through the household's brush to verify the identity card to match; the other is through the subway line personnel bayonet, identify the prisoners in the crowd And suspicious molecules.