Smart Engines Smart Endzhins

Smart Engines was founded in 2010 by scientists from a number of institutes of the Russian Academy of Sciences in the field of image recognition and processing and video stream commissioned by large IT companies in the USA and Europe. The company employs more than 45 employees from Russia, among them 16 people have academic degrees.

In February 2015, Smart Engines, a company specializing in the creation of vision technologies in static and dynamics, announced its entry into the Russian market.

The main areas of research and development of Smart Engines are:

• Image recognition and processing;
• Machine Learning and Algorithmic Optimization
• Recognition of text information in the video stream;
• Mobile solutions in the field of image processing and  document recognition;
• Create high-load systems.

The strategic customer and partner of Smart Engines is Jumio, a developer of identification tools for payment systems. Smart Engines document recognition solutions are used in social networks, airlines, banks, insurance companies and e-government systems.

Embedded Smart Engines technologies can be used to recognize data from various identity cards, bank card data, car numbers, etc., in a video stream. In this case, the source of the video stream can be a camera of a mobile device or a regular webcam.

Smart Engines solutions make user interaction as interactive as possible, as all image processing and recognition takes place on a mobile device in real time. Unlike other solutions in which recognition is carried out exclusively on the server or in the cloud, the Smart Engines approach allows you to ensure the security of payment and personal data processing.

The key Smart Engines technologies are:

• Smart IDReader - technology for recognizing authentication documents;
• Smart PassportReader - data recognition technology of the passport of the Russian Federation;
• Smart CardReader - technology for recognizing bank card data;
• Smart MRZReader - technology for recognizing a machine-readable document area (MRZ line);
• Smart PDF/A - technology for compressing electronic images of scanned or photographed documents;
• Smart ReSquare is a technology for highlighting boundaries and normalizing a document in a video stream.

Based on these technologies, programs have been developed for mobile devices, desktop and server applications for end users.

Technology software developers are available as  an API as part of the SDK for iOS, Android, Windows, Linux, Mac OS, Elbrus. The SDK also includes documentation and integration examples.

Demo applications for recognizing the passport of the Russian Federation, MRZ lines and IKEA labels for mobile devices running iOS and Android are available for download  in the App Store  and Google Play.

Performance indicators

2023:15% Increase in Revenue from Document Authentication Technology Stack

Smart Engines increased revenue by 15% in the fight against fraud. The company announced this on January 30, 2024.

50% of the growth in revenue of the Russian company Smart Engines was provided by revenues from the sales of the technological stack for verifying the authenticity of documents to banks and fintechs. The anti-fraud system based on AI, protected by a dozen patents, was presented only in the second half of the year. In total, in 2023, Smart Engines earned over 460 million on passport and document recognition systems, which is 15% more than a year earlier.

In 2023, Smart Engines introduced two lines of authentication software solutions. The first - the document authentication system using scanners - is designed to verify the physical authenticity of documents of the Russian Federation and the CIS countries. It was developed with the financial support of the Innovation Assistance Fund. The second detects forgeries of documents in digital channels. The project was implemented with the involvement of funds from the Russian Information Technology Development Fund (RFRIT) within the framework of the Digital Economy national project.

Fraudsters, seeing the process of checking documents using the Smart Engines anti-fraud system, simply unfold and leave, - said one of the clients.

Reliable protection and multiplication of the depositor's funds is probably the main responsibility of any bank. For the customer, reliability is fundamental when choosing a bank. That is why banks make every effort to improve security systems. Our technologies help to identify and prevent cases of fraud in digital channels and during service in branches, reducing losses of banks and improving the quality of omnichannel service, - said CEO of Smart Engines, Doctor of Technical Sciences Vladimir Arlazarov.

A quarter of the proceeds were provided by sales of software for recognizing QR codes, bank cards and phone numbers through a web browser.

Interest in WebAssembly technology, which allows you to port the usual functionality to the online version of banks, jumped sharply in 2022, when the applications of Russian banks that fell under sanctions were removed from the App Store and Google Play. These banks faced difficulties in providing services in mobile applications that could not be updated in the usual way. Applications that fall under the restrictions of banks cannot be installed on new iPhones.

In 2023, the Smart Engines team introduced a proprietary technology - UniversalPay. The system greatly simplifies payments in a banking application or Internet bank, since it combines several important functions - from reading QR to recognizing phone numbers and bank details.

In addition, as part of a partnership with VTB, the company has expanded the functionality of the Smart Code Engine software product. Now users can instantly recognize all payment details - TIN, BIC, PPC and others.

Smart Engines Key Business Results 2023:

• Revenue exceeded 460 million rubles (+ 15% YOY);
• A record number of QR payments - 50 million per month - using Smart Engines software;
• New technologies are presented - for recognizing the registration stamp with handwritten filling, cropped QR and 16 more developments;
• 2 new products have been released for document authentication;
• Clients - 10 out of 13 systemically important banks of the Russian Federation;
• The MTS ecosystem, a long-term Smart Engines client, became the company's strategic partner in the field of product promotion and implemented the first sales of AI solutions.

2022: Revenue growth by 43% to RUB 400 mln

Sales of software for recognizing QR codes, bank cards and phone numbers through a web browser brought Smart Engines 37 million rubles in 2022. Revenues rose from zero to 9% in total revenue exceeding 400 million rubles. The software was bought by banks whose apps were removed from the App Store and Google Play due to sanctions, as well as other companies. Smart Engines announced this on January 23, 2023.

Smart Engines sold 216 million recognition worldwide in 2022. Compared to 2021, Smart Engines revenue increased by 43%, and the number of recognition sold - by 91%.

Smart Engines introduced browser software in June 2021. It uses WebAssembly (WASM) technology. It is supported by Mozilla Firefox, Google Chrome, Chromium, Microsoft Edge and Apple Safari.

Interest in WebAssembly jumped sharply in 2022, when the applications of Russian banks that fell under sanctions were removed from the App Store and Google Play. These banks faced difficulties in providing services in mobile applications that could not be updated in the usual way. Applications that fall under the restrictions of banks cannot be installed on the latest iPhones.

Foreign companies are interested in the browser solution for recognizing bank cards from Smart Engines in January 2023.

The growth of our business was due to the development of the functionality of Smart Engines products and the attraction of new customers, primarily in Russia. The demand for recognition of barcodes, bank cards and phone numbers in the browser was further stimulated by external restrictions. I believe that this trend will continue in 2023, and we will see revenue growth by at least double-digit indicators in percentage terms, "said Vladimir Arlazarov, CEO of Smart Engines, Ph.D.

Algorithms Smart Engines using proprietary GreenOCR technology do not smartphone allow you to recognize your passport RUSSIAN FEDERATION in 0.15 seconds. For each recognition, 0, 0001432891146 grams are allocated. CO ₂

We calculated that planting 1 hectare of oak forest will compensate for all recognition forever. In October 2022, in the Ugra National Park in the Kaluga Region, we planted 4,000 seedlings of petiole oak. This oak forest is capable of absorbing 556.7 tons of carbon in 50 years, - said Vladimir Arlazarov.

History

2024: Discovering a way to improve the efficiency of neural networks

Smart Engines scientists have found a way to increase the efficiency of neural networks. The method is based on a fundamentally new quantization scheme, thanks to which the speed of work increases by 40%. Smart Engines announced this on April 25, 2024. Read more here.

2023

41 scientific papers and 25 patents per year

Smart Engines reported on scientific results for 2023. In 2023, the company's researchers published 41 scientific papers in international and Russian publications, received 25 patents, registered 10 software components and made a number of scientific discoveries. The company announced this on March 22, 2024.

Of the 34 scientific articles, 4 were published in the journals of the first quartile (Q1) Web of Science (WoS), 11 articles in the journals of level Q1 based on Scopus.

Two employees of the company defended dissertations for the degree of Doctor of Technical Sciences, and two more for the degree of Candidate of Technical Sciences.

Four Smart Engines specialists teaching at the Department of Cognitive Technology MIPT (based on Smart Engines) were awarded the title of Associate Professors.

In August 2023, Smart Engines researchers unveiled their first public annotated body (datacet) of MIDV-Holo hologram document images. With MIDV-Holo, developers from all over AI algorithms attacks the world can train their best to fight presentation attack and rebroadcast attack attacks, which are carried out by scammers when they photoshop passports.

One of the scientific results of the 2023 Smart Engines was presented in an article published in the journal Expert Systems with Applications (Q1 WoS). It presented a protocol for automated neural network detection of coronavirus by CT, in which the patient's X-ray dose was reduced without a significant impact on the quality of diagnosis.

In April 2023, the journal Mathematics (Q1 WoS) published a study by Smart Engines employees, in which they presented a poneuron approach to training quantized neural small-bit networks. Such neural networks work several times faster than classical analogues, and the proposed approach is easy to implement and allows you to achieve the same quality as more complex methods.

Smart Engines has also developed a universal architecture for robustly evaluating PESAC (Parallel Effective Sample Consensus) hypotheses, which allows you to implement a wide class of algorithms and accelerate the localization and tracing of documents by 2-3 times (relative to other RANSAC architectures) on central processors due to efficient data organization and parallel processing. This result was published in the interdisciplinary journal IEEE Access (Q1 Scopus).

The company's scientists also presented work in 2023 on rapid detection of the machine-readable area (MRZ) for document recognition on mobile devices. The detection algorithm presented by the researchers is based on the author's architecture of the neural network, which finds MRZ in just 16 milliseconds on the iPhone SE 2020. In other words, on a 2019 mobile processor, the detection rate is 62 frames per second. The technology described in the article is already used to recognize MRZ in the Smart ID Engine system, which allows you to enter passport data 20 times faster and 2 times more accurate than a qualified operator.

In 2023, the Russian research company Smart Engines received 25 patents - this figure was a record in the developer's history. The patents were based on the inventions of 70 employees involved in the development of technological solutions for recognizing and verifying the authenticity of documents, bank cards, barcodes.

In 2023, Smart Engines scientists received patents in the United States for six inventions at once. The company has patented a number of its key company technologies - a method for identifying documents, a method for integrating personnel in a video stream, a method for stopping text recognition in a video stream, and a method for using the Hough transform in networks. The company received another patent for a key invention in the field of computed tomography. This is an absolute record for a Russian developer, since before that the company had only one American patent received in 2019. Patented solutions are already used in Smart Engines software products for autonomous recognition of passports, ID cards and other documents.

The company's projects were implemented with financial support from the Russian Information Technology Development Fund (RFRIT) within the framework of the Digital Economy of Russia national project and the Innovation Assistance Fund.

The development of fundamental science and technological potential are the most important components of our company's success strategy. We believe that the development of technologies is impossible without the development of the scientific base, without the development of our own scientific school. I must say that 2023 was a success both from the point of view of the company's business, we received record revenue and profit, and from a scientific point of view. Our employees showed themselves to be excellent researchers and inventors and again exceeded our expectations!, - said the CEO of Smart Engines, Doctor of Technical Sciences Vladimir Arlazarov.

Receipt of 12 patents of the Russian Federation for recognition and authentication technologies for the year

The result of the scientific work of Smart Engines, the leader in the document recognition market, was 12 the Russian patents for inventions. This time, the company patented one of a dozen of its technologies used to authenticate documents. This development - combined with other types of checks - is used banks financial by organizations, notaries, operators communications and other companies to prevent fraud. The attacks company announced this on December 14, 2023.

As follows from the description of the invention to the latest patent, scientists have trained AI to determine what is depicted in the presented photo: the document itself or a black and white copy. Proprietary algorithms compare the color histogram of the document shown in the photo with the histogram of a template of this type. If they do not coincide, the document is most likely a black and white copy. This method is resistant to light drops, which is extremely important when analyzing images of documents in conditions of uncontrolled shooting.

Earlier, Smart Engines noted a series of other achievements. So, she created and patented a system for multispectral authentication of documents. It instantly extracts data from passports of citizens RUSSIAN FEDERATION and others, countries CIS including,,,, Armenia,, and Belarus Kazakhstan Moldova Tajikistan Uzbekistan checks the document for authenticity and validity in three ranges - optical, ultraviolet and infrared. It takes no more than three seconds for the Smart Engines system to complete the input and verification process.

In addition, the company received patents for other key inventions in the field of document recognition and authentication:

• System of multi-level identification and identification procedure using optical recognition of submitted documents;
• A method for detecting fluorescent fibers on images of identity documents in the UV spectrum;
• A method for neural network control of text data on document images;
• User Identity Document Authentication System;
• A method for detecting and localizing a falsified area in a JPEG image;
• Automatic scanner with a multispectral authentication system for passport and visa documents;
• A system for identifying a person's identity from a face image with the issuance of documentary evidence;
• and other developments.

{{quote 'Our company is founded by leading scientists and inventors of Russia in the field of AI. And the whole business is permeated with a spirit of creativity and invention. Using a rigorous scientific approach to solve problems of analysis and processing of documents, we try to achieve perfection of functions of our products. And this year was very fruitful, we received 12 patents of the Russian Federation for key functions of products. In many ways, we owe this result to our investments in our scientific school, which not only allows us to make the best products on the market, but also to educate excellent scientists and inventors, - said the CEO of Smart Engines, Doctor of Technical Sciences Vladimir Arlazarov. }}

Smart Engines patents its developments abroad as well. In 2023, the company received five American patents for key document recognition technologies in the video stream. In total, the Russian developer already has 6 patents in the field of AI in the United States as of December 2023.

Patented AI-based developments have already been introduced into all of the company's software products.

TAdviser interview with CEO Vladimir Arlazarov

In December 2023, Smart Engines CEO, Doctor of Technical Sciences and Associate Professor Vladimir Arlazarov spoke in an interview with TAdviser about the factors that most strongly drive the development of OCR technologies. Read more here.

Obtaining a patent for AI "stop" technology in the USA

In November 2023, the Russian company Smart Engines announced the receipt of another patent in the United States. This time we are talking about technology that allows you to stop the work of artificial intelligence (AI). First of all, it will be useful when recognizing text on video.

According to the press service of Smart Engines, the patented system is able to predict the result of recognition of the next frame and, if it does not allow improving the current recognition result, then a decision is made not to continue processing. The invention of Russian specialists, according to them, helps to speed up the decision to stop AI tenfold. Thanks to this, the process of recognizing documents on mobile devices is faster and more accurate. In addition, the development allows you not to waste unnecessary computing resources and save battery power on the device, Smart Engines said in a statement.

By November 2023, the patented technology is already being used in all Smart Engines software products. With its help, more than 50 million QR codes and bank cards for making payments, millions of passports and other documents in the processes of remote and field service are read every month in Russia.

It is clarified that in January-November 2023, Smart Engines received five patents in the United States. Previously patented:

• document identification method;
• a method of integrating personnel in a video stream;
• a method of stopping text recognition in the video stream and a method of using the Hough transform in networks;
• a key invention in the field of computed tomography (which one, the press service of the company does not specify).

Smart Engines also receives patents in Russia: in January-November 2023, the company was issued 15 patents for inventions in the Russian Federation.

Achieving carbon neutrality

Smart Engines on June 16, 2023 announced that it had achieved carbon neutrality. The company managed to fully compensate for the carbon footprint by participating in forest restoration projects in Russia.

In May 2023, Smart Engines employees planted 3.5 thousand cherry oak seedlings on the territory of the Smolenskoye Poozerye National Park, located in the Smolensk Region and protected by UNESCO.

Scalloped oak is listed in the Red Book of the International Union for Conservation of Nature and is among the trees growing in Russia one of the leaders in its ability to absorb carbon dioxide, second only to aspen and birch. A hectare of oak forest, according to the forest restoration service "Plant the Forest," is capable of absorbing 212 tons of carbon in 50 years.

One recognition carried out using Smart Engines AI technologies requires energy equivalent to 0.0001 g ^{of CO2}. The planting of 1 hectare of oak forest, the company calculated, compensates for the environmental damage from all recognition for a foreseeable prospect.

This is the second reforestation initiative supported by Smart Engines. In the fall of 2022, employees of the company planted four thousand seedlings of petiole oak in the territory of the Ugra National Park in the Kaluga Region.

Forest restoration activities are part of the ESG strategy that Smart Engines formed in 2020. It provides for a consistent reduction in energy consumption and compensation of the carbon footprint. In the same year, Smart Engines joined the United Nations Global Compact (UN).

All products of the company use the author's nature-saving GreenOCR character recognition technology. It provides high quality and speed recognition with minimal energy consumption and environmental impact.

GreenOCR is being developed in accordance with the doctrine of artificial intelligence. It is based on three basic principles: sustainability, security and privacy.

TAdviser interview with CEO Vladimir Arlazarov

Modern companies and organizations are rapidly turning into digital organizations and translating interactions with their customers, partners and employees into electronic form. In these operations, people are increasingly involved remotely, which requires confirmation of the person's identity. In parallel with the growth of digital interactions, the number of digital crimes committed using forged documents is also growing. About how criminal methods of this kind evolve, and advanced Russian technologies help companies fight them, TAdviser in June 2023 told Vladimir Arlazarov, CEO of Smart Engines, candidate of technical sciences. Read more here.

2022

Penalty for illegal use of Russian software

At the end of November 2022, it became known that Citibank was fined for illegal use of Russian software. This is a Smart IDReader product designed to recognize identity documents. Read more here.

Smart Engines experts taught AI to recognize multi-layered microstructures of the human brain

On August 8, 2022, Smart Engines announced that, together with colleagues from Russia, Italy and Germany, for the first time, they automatically segmented a 3D image of a human olfactory bulb using neural networks. Read more here.

2021

Agreement with CNR-Nanotec to create an automated tomographic data processing system

The Italian Institute of Nanotechnology CNR-Nanotec has signed a cooperation agreement with the Russian scientific company Smart Engines for 3 years to implement a project to create a new automatic tomographic reconstruction system. Smart Engines announced this on December 17, 2021. Read more here.

Using Smart Engines AI Algorithms in Passport Recognition Software

On April 2, 2021, it became known that the software developments of the Russian company Smart Engines are used to recognize the passports of tourists arriving in the Seychelles. Read more here.

2020

Accession to the United Nations Global Compact

Smart Engines was the first the Russian development company artificial intelligence to join the United Nations Global Compact (). UN This became known on October 15, 2020.

The Global Compact is an international initiative aimed at involving business structures in sustainable development activities and forming a global partnership between the public and corporate sectors. Over 20 years, more than 15 thousand companies from all over the world have joined the global treaty. In Russia, 74 organizations are parties to the agreement on October 15, 2020. By joining the UN Global Compact, the company reaffirms its commitment to 10 principles of sustainable development in the fields of human rights, labor relations, environmental protection and the fight against corruption.

The approval of our entry into the global compact is both a declaration and confirmation that Smart Engines activities are conducted in accordance with modern goals adopted in the international community. In our work, we are focused on taking care of the environment and minimizing the risks associated with the use of technology. This is the beginning of a new path and public activities in the field of sustainable development at a high international level, where joining the UN State Duma can be considered a starting point, "said Vladimir Arlazarov, Director General of Smart Engines, Candidate of Technical Sciences. - The highest quality standards, environmental responsibility and personal data security is what our GreenOCR approach is based on. In the market for solutions using artificial intelligence and recognition systems, we are the conductors and propagandists of the ideas of responsible technology development and responsible consumption. Our activities are aimed at creating effective software and neural network solutions that reduce the energy consumption of end devices and reduce carbon dioxide emissions into the atmosphere. Also in our developments, we are focused on minimizing the risk of personal data leaks and preventing privacy violations[1].

Agreement with Tessi to promote joint solutions in the European market

On March 10, 2020, it became known that the company, Tessi one of the leaders, French IT market entered into a technological partnership agreement with Skolkovo a member of Smart Engines, a developer. Smart IDReader More. here

Skolkovo Resident Status

Smart Engines Russian , a developer in the field of recognitions document systems based on artificial intelligence , announced on March 6, 2020 that it had become a resident. " Skolkovo Resident status will allow the company to receive tax benefits, administrative, organizational and communication support for the fund. 

Smart Engines is primarily a scientific organization: we have more than 60 researchers on staff, of which 18 have academic degrees. Our scientists are working on research into new neural network architectures of the HoughNet family and a mathematical model of a bipolar neuron. Smart Engines products are the result of commercialization of many years of basic and applied scientific research. They are used by clients from more than 20 countries. By the end of 2019, we implemented the first stage of measures to bring our AI developments to the international market and gained important experience. We expect that interaction with the Skolkovo team will allow us to use new promotion tools aimed at a wide range of foreign customers in the b2b segment, as well as increase investments in R&D, - said the CEO of Smart Engines, Ph.D., Vladimir Arlazarov.

2021: Turnover growth to €3.08 million

Russian The developer of green AI Smart Engines technologies on January 31, 2022 shared the results of commercial scientific and activities in 2021. In 2021, Smart Engines sold AI for 113 million secure document recognition, solved the latest fundamental challenge in passport recognition RUSSIAN FEDERATION and created CO² neutral AI for OCR. The main U-turn of the passport of the Russian Federation is now recognized in 0.15 seconds on, mobile phone even if it is removed by a "book," throwing out only 0.0001 grams. CO2

Business growth

• The number of recognition sold for safe extraction data in solutions and mobile applications Internet other information systems amounted to 113 million.
• The company's turnover in 2021 is 3.08 million euros.
• The customer base of the Smart Engines product line increased by 26% compared to 2020.

Results of scientific activity

• At the basic department of MIPT and under the guidance of the company's scientists, 10 bachelors, 13 masters were trained, and 2 researchers from the company's employees defended their Ph.D. dissertations;
• Published 26 scientific articles in journals, including journals rated Q1;
• 15 reports presented at international conferences;
• Participation in ICDAR 2021;
• Collaboration with Italian nanotechnology institute CNR-Nanotec.

Faster recognition

• The speed of bank card recognition has been doubled. The frame is now recognized in just 0.035 seconds.
• The speed of recognition of QR codes has increased 1.6 times, the number of unread codes has been reduced by a third.
• Recognition of ID documents accelerated by 40% over the year (for infographics using the US Driving License as an example).
• Due to the 8-bit playback model, all neural networks accelerated by 40%.

Improve recognition accuracy

• Increased accuracy of document recognition: in 2021, the number of errors decreased by 15%.
• Card number recognition accuracy reached 99.68%.

Reduce the size of the recognition library

• In 2021, we made the recognition library even more compact - on average, the integration of Smart ID Engine into a mobile application in a configuration for recognizing a passport of the Russian Federation increases it by no more than 10 megabytes.
• Reduced the volume of neural networks used by 30%.
• Reduce memory and configuration size for Chinese, Japanese, and Korean.

Number of document types

• The updated version of the Smart Code Engine allows you to recognize the bank card data of 21 payment systems in the video stream and in photos.
• Added 316 new document types and 513 new templates.
• We recognize 1856 certifying documents, bank cards and barcodes of 210 jurisdictions.

Changes:

• WebAssembly support.
• Document recognition within a web page.
• Recognize a bank card from a photo and KYC mode for cards (with a partially hidden number).
• Recognition of the passport of the Russian Federation in projective mode on a mobile phone ("book" mode).
• Support for scanning and decoding AAMVA barcodes.
• Recognition of ID cards on both sides within a single recognition session.
• Support for Arab and Persian ligatures.
• Multiple barcode scan mode.

Fraud prevention

Creation of 12 complex artificial intelligence systems for the detection of digital and physical counterfeiting of ID documents: Computation Document Forensic AI, Image Fraud AI, Pixel Fraud AI, Font Fraud AI, Text Fraud AI, Template Fraud AI, Photo Fraud AI, Data Fraud AI, Stamp & Seal FRAUD AI, Liveness AI, MultifRAI, Fossil FRAUD.

Environmental impact

As a member of the international UN Global Compact initiative, the company develops AI technologies in accordance with ESG principles. Smart Engines researchers were able to make AI algorithms almost entirely ^CO2 neutral. The scientific developments in the field of Green AI implemented in the proprietary GreenOCR technology made Smart Engines recognition 2 times more ^CO2 neutral. With AI algorithms, only 15 trees are required to compensate ^{for CO2} emissions from 113 million recognition, because one recognition consumes energy equivalent to 0.0001 gr. CO2.

2019:49% sales growth and 29% customer base expansion

On January 15, 2020, Smart Engines summed up 2019.

Compared to 2018, the company's sales increased by 49%, and the customer base expanded by 29%. At the same time, in 2019, the company's clients were the big three mobile operators (MTS, Beeline and Megafon), and one of the leaders in the banking sector of PJSC Gazprombank. The total number of recognition sold, taking into account international activities in the USA, Asia, the European Union and the CIS, amounted to 51 million.

The priority for the company in 2019 was the development of its own computer vision technologies, combined under the GreenOCR ideology, which allows recognition applications to work faster and more accurately, and also makes them more user-friendly and energy efficient.

Our business is based on three priority principles for us. The first is the security of user data and minimizing the risk of its leakage. This is very important, especially when you work in Russia and other countries where they are very strict about the processing of personal data. The second is to reduce the negative effect on the environment. We emphasize that the technology, if it is really good, should also be 'green'. Third, we help our customers transform into remote services in line with current trends, which makes their business more dynamic and profitable, notes summing up the year, CEO of Smart Engines, Ph.D. Vladimir Arlazarov

In addition to business achievements, the indicators of our scientific activities are very important for us, since we are a scientific team. In 2019, we increased our staff by 40%, inviting strong researchers from among recent students to the team: at our basic department MIPT , we prepared and graduated 17 bachelors and masters. One of our employees successfully defended his thesis; we took part in specialized international conferences ICDAR, ICMV, where we presented 35 reports, comments technical director of Smart Engines, Ph.D. Dmitry Nikolaev

The company says that in 2019, scientific papers were published regarding the development of the HoughNet architecture, designed to more effectively train and execute neural networks when solving various problems in the field computer vision, as well as the results of scientific research devoted to the development of low-potential and bipolar neural networks with increased specific performance and reduced power consumption.

2017

Income from document recognition in Russia - over €1 million

In 2017, Smart Engines earned more than 1 million euros on document recognition in Russia, which is 4 times higher than in 2016.

Most of the income came from the flagship product Smart IDReader, designed to recognize a Russian passport, driver's licenses, bank cards and other documents in mobile applications and Internet solutions.

Over the past year, the company's partners have become, AT Consulting the company Leta and the developer operating system Sailfish Mobile OS RUS company. "Open Mobile Platform" Smart Engines also continued to collaborate with the microprocessor developer "" Elbrus company, MCST system integrator "" and Corus Consulting the developer of solutions for remote. identifications Sum&Substance

Partnership with SMTDP Tech

In early 2017, Russian developers of image processing and deep machine learning technologies, Smart Engines and SMTDP Tech, announced the start of a collaboration that provides for the creation and promotion of a joint solution for recognizing identification documents and identifying counterfeits. The companies are working on the integration of technologies and the creation of a joint product that solves the tasks of document recognition and includes elements of protection against fraud for Russian and international organizations of the financial sector, the air transportation sector and the sharing economy.

The capabilities of the created solution include the functionality of recognizing authentication documents on mobile and server platforms, as well as identifying forgeries of documents in photographs and scanned copies. Its task is to signal to an employee of the company that the received document is fake or that its owner is on the terrorist list or other sanctions list or that the monetary transaction may carry risks for the company where such a client came. It will be available to customers as a standalone product with the ability to integrate it into specialized AML/KYC (Anti-money laundering and Know your client) systems.