With the growing popularity of computer vision and facial recognition, businesses strive to adopt innovations to keep heads above water. As Accenture’s data scientists report, in such areas as security, customer interactions, retail, marketing, etc., they predict the future full of tech invasion, no less.

In this future, different scenarios are possible. It spurs inevitable discussions. That way it was with Amazon’s Recognition when a group of stakeholders expressed worries about potential abuses of the facial recognition technology for surveillance purposes. And not in vain.

Commonly, the main argument for implementing AI-based innovations, e.g., facial recognition, is the intention to increase the ROI. This situation can’t but raise ethical questions. How should organizations tackle the downsides of AI? It’s time to figure things out.

Building AI for a reason, not just because one can

The popularity of AI gained momentum after the conference in 1965 when a scientific community first united to study ways to replicate a human brain and enable machines to undertake activities only humans could do. Since that time, computational power and positive experience in the field of data science has dramatically scaled up.

To date, latest tech stack, AI libraries and frameworks allow programmers to build a broad range of AI-based solutions and applications that fit various business needs. While using some provided tech stack, they may pay insufficient attention to the core principles of development. That is to say, ethical principles. Too many of them can create quality applications, but few ask themselves about the ethical aspects of building something novel.

As one of the world leaders in the field of facial recognition, Microsoft expresses concern about the ethics of AI. They have figured out 6 core principles to help make the development more meaningful.

6 pillars of ethical uses

Sharing his view, Brad Smith, Microsoft, said, “Advanced technology no longer stands apart from society…, our personal and professional lives. This means the potential uses of facial recognition are myriad.” The technology is capable of influencing not only business development strategies but also communications, day-to-day personal life, routine tasks, the reputation of one or another organization on the market as well as its reputation among employees.

Some cases of using facial recognition technology brought to light algorithmic bias and raised doubts about transparency and legitimacy. Let’s take for instance the image recognition algorithms in Google Photosthat uncovered racial biases or the face-scanning system in Orlando International Airportthat raised concern over the accuracy of scanning and the fact that there are no strict regulations on how to work with passengers who were misidentified. Given all this, it seems wise to keep an eye on observing guidelines while creating and using facial recognition systems. And here are 6 fundamental ones:

• Fairness
• Reliability and safety
• Privacy and security
• Inclusiveness
• Transparency
• Accountability

Ethical guidelines for gathering and disposal sensitive data have already seized the attention of adopters of facial recognition. To reinforce the results, it’s crucial that each particular organization perceives regulations as a question of paramount importance.

Playing a fair and transparent game

Innovations seem creepy unless they are well explained to the public at large. On weighing all pros and cons and deploying face recognition software, an organization will most certainly face the issues of justified gathering and using personal data for facial recognition. Sticking to the following principles can help make things go right:

• Inform about the use of the technology in any public area. It can help avoid possible awkward discussions about the legitimacy of keeping visitors uninformed and violation of privacy.
• Be watchful on datasets. A facial recognition system is as smart as a training dataset is inclusive. For instance, comprehensive data can help avoid racial bias. But perfection always lies out of reach. Ideally, it’s better to elaborate a strategy on how to treat people whom facial recognition misidentified. Each organization may apply specific measures to ensure a balanced approach to the problem and create a good image.
• Share information about personal data storage. The work of a facial recognition system can provoke prejudice against further uses of sensitive data. Another responsibility of each organization is to ensure private and secure storage of data. An organization should be ready to explain the reasons for processing personal data, inform about any unintended misuses and have clear evidence that the use of data is fully transparent and legitimate.
• Train and raise awareness. Any new technology brings both social benefits and concerns. But it doesn’t mean having one’s head in the sand. Oliver Schabenberger, SAS, admits, “Even for many sophisticated users, AI still is a black box.” More importantly, he suggests that it is time to illuminate this black box, in other words, observe its work and explain the pros and cons of using it. If everyone within an organization comprehends the principles of using data for face recognition, they will calmly respond to bias. This kind of faith in new technology is indispensable.

The use of any technologies can equally be for good or evil. Artificial intelligence is developing by leaps and bounds to assist humans in transforming the present-day reality to the better. However, meaningless tech development may lead to social misunderstanding and provoke ethical debates. Thinking ahead and keeping regulations in mind — that’s what can help build a brilliant future of facial recognition.

Enjoyed this article? Find more stories here: https://indatalabs.com/blog

Tap into the power of data science!

Ethics of Facial Recognition: How to Make Business Uses Fair and Transparent was originally published in TDS Archive on Medium, where people are continuing the conversation by highlighting and responding to this story.

Teacher’s role in an AI golden age

As Amazon CEO Jeff Bezos said, we are witnessing a renaissance of AI technologies. And if not to get to the heart of the matter, machines taking over from humans in a large proportion of tasks may seem alarming. AI has already reduced the demand for workers doing mundane tasks, improved the standards of leaving and enabled humanity to solve complicated healthcare, logistics, security tasks that require informed decision-making. And that’s all for the better.

AI is not a Pandora’s box. Instead, it is predicted to create more jobs than it is going to eliminate. But many professions, such as teaching, indubitably require the human mind, not an artificial one. Thus, thought leaders, intellectuals, and savvy professionals trained to communicate ideas between humans and machines will become the most sought-after specialists.

When a teacher needs AI assistant

The workload dumped on a teacher often exceeds what is reasonable. A good and passionate teacher is expected to handle up to 30 students within a classroom, make them get down to learning, stop giggling and playing with gadgets, monitor student academic performance, grade homework, prepare lesson plans and on and on. These challenges are daunting, let alone teachers’ salaries. Too exhaustive and often underpaid, the teaching profession is losing popularity, which leads to teacher shortage worldwide. However, the tech revolution is on the way to transform teaching and learning as we know it today.

AI is on the mission to give a helping hand with planning, personalizing, visualizing, and facilitating the learning process. Let’s delve into the following AI-powered innovations that can assist teachers in making not only students but also learning process smarter.

Pushing classroom limitations

Traditional classrooms, blackboards, and manuals have already become a part of a retrograde understanding of the learning process. Tapping into technologies, teachers and students can go beyond the borders of what seemed possible before. Hopefully, soon the only limitation left in the educational area will be the desire of one or another person for self-development.

The use of AI-powered tools enables creating global classrooms. It brings new learning opportunities for students with visual or hearing impairments and speakers of other languages, allows teachers to consider different learning styles (making things easier for visual, auditory learners, etc.) and personalize learning as much as possible to improve academic performance.

To provide an example, a free plug-in for PowerPoint Presentation Translator creates subtitles in the real-time mode. Such technologies as Azure Cognitive Services, AI-powered speech recognition and translation allow students to follow a teacher in a most convenient way.

1. AI fuels Automated Speech Recognition and Natural Language Processing generates the script.

2. Machine Translation adapts the text for learners speaking different languages.

3. Students can listen or read the text displayed on the teacher’s screen or join the conversation via personal gadgets.

The capabilities this solution provides can’t be overstated, since ultimate personalization of learning and its global accessibility is what everybody craves for.

Robots for no learning in isolation

So many students worldwide are unable to attend school due to long-term illnesses. But AI is here to enable smart distant learning opportunities. Robots can replace students in classrooms ensuring real-time presence, even though it is physically impossible, and combating the isolation of a student from classmates and teachers. This technology allows students to get out of their hospital beds and control what is going on around using a tablet. Such robots have a connection with Microsoft Azure IoT Hub providing a student with all the video and audio connections and allowing participation in class discussions.

The educational robots market is growing with the goal to deliver affordable educational hardware and software solutions. However, the demand for this technology is restricted as depends largely on teachers’ readiness to incorporate robots in the learning process.

Teacher’s assistant

Automation of routine activities as part of the digital transformation is a present-day trend. And the educational area is not an exception. How much time teachers spend grading homework and preparing reports? A crazy amount. Would they re-plan their time to take advantage of new opportunities and bring their students to new heights? For sure, if only they had a chance to entrust someone else with tasks requiring little intellectual resources.

Utilizing Microsoft Azure and Machine Learning, such educational software as Gradescope can assist in grading student works or submissions, generating and sharing detailed analytics. An AI assistant can help teachers tailor learning plans according to different student’s needs. For instance, CENTURY Tech gathers data on every student’s learning progress. Analyzing data on each student in particulate, AI takes into account the details a human teacher could fail to notice or pull together. AI is able to provide data-based predictions of student’s success or failure, identify gaps in knowledge, highlight strengths and weaknesses and suggest where extra tutoring would be of help. Omniscience is a real power of AI.

The students of today will live and work in the future where AI will be a daily reality. Time for paper manuals and blackboards has passed: it’s time to expose students to cutting-edge technologies already at school. The modern learning process requires more interactivity and personalization to make knowledge more accessible all over the world regardless of a person’s place of living, language, health impairments, or learning abilities.

AI can easily become a students’ and teachers’ best friend providing personalized tutoring and assistance at the point of need. But what AI can’t do is to be a source of inspiration for learning, unlike a human teacher. The importance of empathy in the learning process is higher than it seems. And that is the main downside of educational AI-powered solutions, and that is why AI isn’t likely to entirely replace human teachers.

Enjoyed this article? Find more stories here: https://indatalabs.com/blog

Tap into the power of data science!

Is AI here to replace human teachers or is it a teacher’s assistant? was originally published in TDS Archive on Medium, where people are continuing the conversation by highlighting and responding to this story.

What brings all technology-based innovations together? It is the developers’ craving for making them work in the likeness of human beings. This desire spurs the evolution of AI and Deep Learning — technologies that are going to affect a number of related jobs.

As Gartner claims, today before one innovation is implemented, two others arrive. But the outcomes are controversial. For instance, AI is predicted to eliminate more jobs than it creates through 2019, while the technology is also believed to create enough new jobs and enrich current people’s careers in terms of effectiveness and productivity. In the future, among the specialists whose jobs will be eliminated may be template designers and GUI programmers. And the reason for that will be the accelerated growth of “human-like” programs, namely Artificial Neural Networks.

RNN architecture and the way to use it in automatic programming

Automated processes are gradually implemented to boost the effectiveness of intellectual tasks, e.g., generating source code. There is nothing sort of miracle here: the code is automatically generated by a written program based on certain functions. For instance, an open-source Google/MIT App Inventor allows users to drag and drop some needed functions in the development area, connect them to each other, define the way an application is going to work, and get the generated source code based on the prepared template. Using this application, even newcomers to programming make little efforts to create Android-based apps.

But what about bringing coding to a whole new level? Everybody would agree creating GUI is a time- and effort-consuming process. Working on small-scale projects, developers are often in charge of creating GUI, although it is not among their priority tasks. What if they could get rid of this burden? What if a program could capture and recognize screenshots of a GUI and generate the source code from image data?

There is something magical about Recurrent Neural Networks (RNNs). In fact, to train a multi-layer RRN is not so difficult as it seems. This type of neural network architecture is organized into recurrent layers: the input layer, hidden layers, and the output layer.

Recurrent layers allow feeding the information from previous time steps and combine it with the input of the current time steps. It means that the order of input information matters. This is the reason coders often employ the algorithm of gradient descent to training a neural network. The algorithm allows continuous improvement of the desired output, thus making a network “more intelligent” with every new update.

RNNs can be successfully used in many fields, where data can be represented as a sequence, i.e., for pattern recognition and automatic code generation.

Pix2Code and generating GUI code “from sketch”

Here we get to the main issue: how to leverage AI and deep learning to facilitate the process of coding based on designers’ mock-ups? To date, there are hosts of programming languages specific to different systems where custom software is supposed to run. It makes the process of implementing GUI code tedious and time-consuming.

UIzard Technologies has presented a Pix2Code based on several architectures of neural networks, including RNN, to enable generating the source code from a screenshot. The model was trained on a small dataset, and therefore, Pix2Code is just the beginning of automating the process of code generation.

The developers came quite close to perfection, addressing the following three issues:

Computer vision. Machines cannot recognize and process objects or characters on input images. The problem was solved by means of Convolutional Neural Networks (CNN) specifically trained on image data.

Language problem. RNNs, actually responsible for text recognition and code generation, suffer from vanishing and exploding gradient problem. To eliminate this problem, they used another neural architecture, Long Short-Term Memory (LSTM).

Network training issues. The training process involved experiments since there was no ready-made image-code dataset for the network. However, the network demonstrated sufficient capacities in linking texts, images, and codes and generating a 77 percent accurate code for various platforms.

The generated code doesn’t require significant changes, and it is compatible with different platforms, like iOS, Android, and Web. As it was mentioned, the current accuracy of the code is high enough. But before turning into a great tool, Pix2Code should be significantly improved through training on a large database.

Perspectives

The unity of AI and Deep Learning is in line with the trend of combining design and coding in a single process. Nonetheless, the perspectives of moving in this direction are controversial.

On the one hand, the development process will be drastically accelerated, which will allow developers to give more priority to other project tasks. However, developers won’t adopt the fully automated process of GUI code generation until the number of possible user interfaces is limited to the number of screenshots that a neural network is trained to recognize. In the future, this issue can be solved by training a neural network on a large-scale dataset.

What seems more challenging is the task of training a network to single out the algorithms of GUI code generation. It means such NN shouldn’t just make a linear choice based on familiar patterns. It is supposed to “think” and to process data the way a natural neural network does.

On the other hand, the automation of code generation process will undoubtedly affect the number of related jobs. It is reckless to say that there will be no need for such specialists, but the AI leap forward will surely bring significant changes to the job market.

Enjoyed this article? Find more stories here: https://indatalabs.com/blog

Tap into the power of data science!

How Artificial Neural Networks Can Code Smarter Than GUI Programmer was originally published in HackerNoon.com on Medium, where people are continuing the conversation by highlighting and responding to this story.

What does the mobile presence mean for a company in the light of modern marketing realities?

It provides either market leaders or startups with the pool of possibilities. As the mobile rush progresses, a growing number of services providers acknowledge the leading role of the mobile experience compared to the desktop one. Should a website risk not being mobile-friendly, any hopes on customer enthusiasm and loyalty will be futile.

Whilst a fruitful mobile app constitutes an inalienable part of a client’s business, the overall number of common apps available for download is soaring (and most often, apps feature poor UI/UX created to bore users to death). According to the previous year’s stats, Android users had a plentiful choice of 2.8 million apps and iOS owners were slightly behind with 2.2 million ones in the Apple App Store.

So, when clients reach out to a software vendor with a request to design and develop another custom mobile business app, they often have little to no idea of what they really need.

Before making the decision on investing in your mobile app development, ask a business analyst the following 7 questions and figure out the true usefulness of a would-be app.

1. How often will customers run your app?

Mobile devices owners have dozens of apps on board but use only a handful of them. Their favorite list encompasses most frequently run apps with the rest disused and removed. Revise the functionalities of your would-be one and make sure it will be used far more often than a couple times a month.

2. Will some unique features make your app stand apart?

Clients often fail to clearly articulate app’s special or unique features opposite to typical website functions. They honestly but unreasonably believe it is enough to create a client-app that will bring users to the corporate website — and they’re golden. Actually, end-users follow the line of least resistance: since it is easier to log in to the website, downloading a mobile app is a time waste.

A sought-after mobile app should entail some considerable advantages, i.e. better performance, higher usability, more secure payments, etc. Maybe it’s time to give your app another thought?

3. Are your business processes automated enough?

A mobile app may perfectly augment your automated business workflow, or it may inject confusion in the work of staff and become a headache for consumers. Check first the level of your business processes automation: are all the online orders handled as quickly as incoming phone calls or is it possible to book a table online at your restaurant in no time and by using updated booking info? If not, get back to automating business processes and then move on to the app development stage.

4. Will your app support offline mode?

There is a perfect indicator to help you make up your mind on whether your business needs a mobile app or not. It is a straight answer to the question: will this app be helpful to end-users in the offline mode? If they won’t be able to use it offline, you have to seriously think the issue over. Perhaps, the development of a mobile app doesn’t answer your current business goals.

5. Do you have enough market coverage?

It is mistakenly believed that a mobile app itself will win claps from the target audience. In fact, your mobile app can get lost among thousands of similar ones. You’d better take some preventive measures.

First, ensure that your customers are aware of your app. The app is your specific promotional channel, but the percentage of end-users employing mobile apps is fairly small compared to the entire customer base.

Second, make the app adorable and urge customers to learn about its full functionalities, to download it and adopt all the advantages.

Third, avoid creating uninformative titles, cliché descriptions and any tech snags in finding the right app. You need to get it listed in iTunes and Play Market. But the final say on your mobile app’s “to be or not to be” dilemma is after the reviewing teams. Give your product a chance to enter the market.

By successfully promoting your product, you can expect some tangible benefits in return and, in the long run, dramatically change customers’ attitude towards your company.

6. Are you going to multiply the lines of “zombie apps”?

The rise of the “zombies” — apps invisible to consumers — is preconditioned by the fast-changing nature of hardware requirements are constant upgrades of operating systems.

To make the app compatible with top-notch mobile devices, invest more in testing and upgrade services. Otherwise, your app is doomed to join the lines of “zombies” that look odd with their legacy usability issues. Therefore, it is wise to consider the total costs of the development, support, feedback management, compatibility testing plus additional expenses to maintain app’s seamless functioning. You can negotiate software maintenance, support & update costs beforehand to include them in the project budgeting plan.

7. Your app business functions are really that specific and unique?

Don’t dive into the app development without having a glimpse of already available apps. Perhaps, you will be able to find a good fit for your business. Among the most common apps, that crowd the market, are accounting and inventory mobile software solutions, messaging and note-keeping apps and more.

If you need a service app for hotel reservations, restaurant booking, rental services, etc., get your app indexed and ranked in such search engines as TripAdvisor, Airbnb and the like. You will gain access to their client databases and will allow your customers to download robust and user-friendly apps.

What’s more, some communication and document sharing issues may be resolved with the help of popular instant messengers, social networks and cloud-based apps by software giants, such as Google or Microsoft.

All in all, hope on you arriving at some useful answers. The article is not an intent to discourage you from investing in a custom-tailored mobile app. But you are highly recommended to frame the right questions to your BA. It will assist in the elaborating all the phases of app development lifecycle and accurate calculation of budget expenses.

7 Questions to Ask Before Building a Business Mobile App was originally published in HackerNoon.com on Medium, where people are continuing the conversation by highlighting and responding to this story.

Going forward, I leave behind brackets all the well-known methods of verifying the authorship. The article’s main goal is to help website owners make their content secure.

Who needs to steal?

For over 10 years of work, OCSICO’s marketing team has come across a number of web portals filled with the content stolen from us. Most of those websites were removed through our efforts. Due to such kind of activity, we managed to maintain high trust level of our website. Now I can uncover information about those engaged in stealing content. So, who are they?

• Unscrupulous web developers. Almost about half of the unscrupulous web portals we came across were the clones of our corporate website. Basically, they were small web-agencies that simply tricked their clients who had opted for their web development services. Some of their HTML coders used our content as the default filling of their web templates. Others cared to change a corporate logo but stole UX design, portfolio screenshots, product and project titles.
• Dark SEO masters. However naïve it may sound to you, some so-called SEO specialists honestly believe it is enough to copy the content of a high ranked website to gain that equal trust, traffic and ranks. Actually, it doesn’t work but may affect the original website. Therefore, it is crucial to prevent the activities for creating clone websites.
• Lazy copywriters. The temptation of stealing a whole article or several neat paragraphs is often irresistible. Otherwise, someone will have to switch brain on.

Why is shielding content necessary?

First comes SEO reasons. Search engines could often index clone websites much faster than your own original web portal. In this case, your content will work for competitors rather than for you. It may even trigger the rating indexes’ fall, which is absolutely unacceptable for your website.

Secondly, quality content is your investment into the website. And you need is to keep your investments safe, or the invested money will work not on your side but on dishonest competitors.

Thirdly, unique content boosts sales. If prospective clients lose track of your website among many similar others, you will lose orders.

Technical means of content protection

1. Restrict access to websites from an insecure location. It is an effective but controversial technique. Website lock-down is a rare case. Alternatively, you can go less radical way and forcibly redirect users to regional copies of a website.

2. Ban copying and disable context menu by means of JavaScript. This really works for nubby users only. Moreover, not all the browsers support these features. It ends up with that you won’t protect content from skilled content thieves but will significantly reduce your website usability and compatibility.

3. Prevent text selection in CSS-styles. It is not a universal solution too as there is a number of workarounds to copy certain content.

4. Use specific render engines. Maybe you still remember flash-based websites where it was impossible to copy a text in case the author was against it. The engines that allow generating bitmaps instead of native text work the same way. Today, with no system constraints, you can easily employ this technique. But you shouldn’t forget that in this case, search engines will ignore your website and won’t index it.

5. Personalization. This method really works! You will need more content compared to a traditional website and a special CMS that tracks user profiles and offers the most relevant content in separate blocks. Combine the use of the CMS with A&B testing and you will increase the variety of displayed content and will prevent stealing the whole of it.

Structural means

You need to deepen content integration into your website structure. It will make copying of selected pages a pointless activity for the thieves.

Dynamic website structure

Employ specific approaches to organize the content and structure of your web portal. You can replace voluminous articles with the snapshots of topics. Or you can upload new content in separate frameworks or clickable spoilers. End-users will surely appreciate the high usability of a website that adds content on demand. As a bonus, content thieves won’t be able to restore a webpage structure, since they aren’t among your target audience and don’t follow the logic of content presentation. On the other hand, you should be careful with splitting content into frames. A part of this information may slip away from search crawlers and, as the result, affect indexing. You need to work out the architecture of your web solution or even give a thought to a custom web application. This way, you hit two birds with one stone: maintain content attractiveness for search engines and make your website really user-friendly and easy-to-use for visitors and your copywriters.

Copywriting means

Content branding

If a rival copywriter realizes that it is necessary to rewrite more than 30% of the content, the idea of copying your pages won’t seem so attractive any longer. To prevent stealing content, follow these tips:

· Opt for a brand-centered approach to creating content

· Often mention a company name and products

· Make active use of interlinking inner pages

· Insert names of your employees

· Mention office and store addresses

From author’s perspective

When appropriate, write the content in a first person. Insert as many details and references to personal experience as you can. A bit sense of humor and sarcastic remarks will contribute a lot to your article. Several bold author techniques will most certainly scare off unscrupulous copywriters who are always searching for neutral content.

Dynamic blocks

Add such blocks as News, Insights, Recommended articles, etc., on each page. For your target audience, it will be a usability feature, whilst for content hunters, it will be a real nightmare.

Shield Your Website from Content Thieves was originally published in HackerNoon.com on Medium, where people are continuing the conversation by highlighting and responding to this story.