The world of cryptocurrencies is like an ever-changing space that is highly unpredictable and experimental in terms we can not even think of. Recently, a blockchain-based solution provider for businesses and public institutions, Waves Enterprises, took the digital arena in a blow when it announced the launch of a proof-of-concept for face recognition on the blockchain network. The project is operated using Docker smart contracts and a neutral network, which undoubtedly is a more secure, potent, reliable, and decentralized version of the existing solution.
#WavesEnterprise has created a proof-of-concept for face recognition on the blockchain, using Docker smart contracts and a neural network – a more secure, reliable and decentralized version of an existing system!https://t.co/O04geYB0FV
— Waves Enterprise (@wvsenterprise) October 10, 2019
This article is all about the implications of the smart contract technology approach and its undeniably strong usage in face recognition on the blockchain.
Solidarity Between centralized and web 3.0 applications
The primary benefits of employing blockchain technology in the concerned business application are:
- Transparency and consistency- In a decentralized solution, all the nodes share the same transaction history. This implies that in case of an alteration of a record, there will be the need of a consensus of the majority of nodes and also the rewriting of multiple blocks. This is nearly impossible for an attacker to achieve easily, thereby allowing better transparency and reliability of data.
- High levels of security- A particular transaction record can only be appended to the existing network on the blockchain after getting consensus beforehand by a majority of nodes. This restrains an intruding attack on the private blockchain. Also, as blockchain is decentralized, a DoS attack needs major efforts to get successful in corrupting multiple nodes at once. In case of failure, the data can be easily retrieved from the multiple copies stored.
- Proficiency and speed- Blockchain is capable of managing business processes efficiently on its network, thus benefitting the organizations in a big way by fostering a reduction in paperwork, human errors, delivery, and administration expenses.
The current era is all about revolutionary technologies taking over manual business processes in an array of domains. Blockchain technology is indeed a new-age technology that is capable of leveraging business operations to the next level. Let’s understand that with an example of the evolution of renowned terminology, entrance access control.
- At the starting phase, this system required people to use tangible objects like badges, id cards, etc. while undergoing the checking process.
- Later, digital passes came into existence, which sped up the process at a substantial rate. But still there occurred problems as people tend to forget or lose their passes.
- Further, the system evolved into the technology of face recognition, where people would just swap their digital cards for being identified by the system.
As the current system is centralized, it is vulnerable to attacks and infringement by some outside party or from the dishonest party inside the organization. Some issues which occur because of this vulnerability can be explained as follows:
- Suppose there is an accident in a certain of your building, which is only accessible to some trusted employees. However, when you check the attendance logs, it shows that no one was present at the time of the accident at that moment. This shows that the data has been altered by someone, and as the network is centralized in nature, it is nearly impossible to retrieve the deleted data.
- There can also be situations where employees might alter their attendance logs secretly to avoid pay-cuts due to excessive absents. An employee who has been absent quite frequently during the year still manages to have his absences within the permitted number only because he/she has played with the centralized system to turn things in his/her favor.
Such cases are quite common at workplaces operating on a centralized network. It is here when blockchain can rise from the debris and evolve as the strongest remedial measure. We can work on an application that automatically recognizes employees from their images provided by cameras, and record attendance on the blockchain. As blockchains are transparent and immutable, the data remains secure.
Face recognition contract architecture
Wave Enterprises’ face recognition for blockchain solution’s various components are explained below. Have a look:
- A video camera installed for recording the building’s entrance. One can also install multiple cameras in different locations to avoid centralization
- The Enterprise Node which does not need powerful hardware and can be operated on a laptop with internet access for communication with other nodes
- A containerized smart contract which comprises of a Neutral Network (precisely ResNet) facilitating face recognition
- The frontend application which works as the user interface for the network allows viewing of attendance records, the addition of new employees, as well as fosters interaction with the smart contract and nodes easily
The Working Pedagogy of The Wave Enterprises System
- The live video stream from the camera gets transferred to the Frontend Application so that the security guards can view it directly, if necessary.
- The video stream is then broken down into images which then travel to the Node as a transaction trail
- The Node ports the transactions to the smart contract
- As soon as the smart contract identifies a face, it records on the blockchain about the entry of the employee, his personal ID, and a timestamp. There is also kept a record of unknown faces to keep track of strangers or in the case of neutral network failures
- The Node stores the transactions, calculates the daily and monthly statistics, and sends them to the network. The Frontend Application displays the required information.
Some tit-bits about the solution
- As the data needs to be recorded on the blockchain, the speed of the solution is a bit slow but is not something that is substantially noticeable in everyday use. There is a minor lag in the attendance list which in the worst case is likely to take 15 seconds for getting updated
- One can store as much data as the machine running the node can hold. The firm has tested the application with up to 100,000 faces without having any major issues.
- The quality of recognition is independent of the use of blockchain. In case the solution fails to identify the face, it stores the image as unrecognized.