Categories
Tech

This Music Startup is Using AI to Identify Musicians and Hidden Talent

A Los Angeles-based music label start-up, Snafu Records, has decided to use artificial intelligence (AI) to find out hidden talents in the musical industry. The company’s founder and chief executive, Ankit Desai, stated that his firm is the first one to deploy the use of AI, enabling the record label to spot unrecognisable talents who may not have been published, or remain in niche segments of the industry.

Taking a dig at the proportions of the music industry, Desai added that traditional musical labels are not capable of identifying such talents, and they are “ill-equipped” to handle it. To emphasise on this, Desai’s Snafu Records will use algorithms to search for such artists. He further added, “If there’s some girl in Indonesia whose music the world is dying to hear, they’re never going to get the chance. The bridge to connect her to the world doesn’t exist today. The music business is entrenched in a very old way of working, finding artists through word-of-mouth.”

Desai is optimistic that his approach will help uncover many unheard voices and bring them to the forefront. Reports on the matter have stated that the company will seek access to around 1,50,000 sound tracks, which are available on platforms like YouTube, Spotify, SoundCloud and Tiktok, and then evaluate them. Unlike the traditional music filters, which take four to six weeks to recognise an artist, Snafu’s software is capable of finding “undervalued” artists with the release of their first song.

Categories
Tech

This 1896 Short Film Was Upgraded to 4K by AI, and It Now Looks as if Made Yesterday

To anyone that has studied films, the earliest short films, which were actually more like short video clips, remain a matter of great intrigue. The same applies for the film L’Arrivée d’un train en gare de La Ciotat, or The Arrival of the Train (as it is popularly known in English), which was filmed back in 1896 by the pioneers of cinema, Auguste and Louis Lumiere. While it was a fascinating piece of work back then, it feels difficult to connect with today, and at times even looks comical. (Majorly) improving this, one Denis Shiryaev used some neural network artificial intelligence upscaling algorithms on the footage and turned it into a footage that streams at 4K resolution, 60 frames per second. The result, needless to say, is quite breathtaking. Before we get to the new, improved footage, here’s a look at what the original film was.

As compared to the original footage, which was shot by a primitive camera that could process 16 frames per second of film material, the upscaled film looks like a gleaming piece of interjection in a Quentin Tarantino movie, shot in high contrast, true greyscale composite film on 8K oversampling. Even the characters do not look out of place – in fact, it looks like the perfect setting for a high action period drama set in the Wild, Wild West. Take a look at the improved piece here:

There is, of course, certain issues with the footage, which gradual improvements to AI and neural networks will almost certainly rectify in future. Barring the few interpolations and video artifacts left in the footage, the footage looks like it could have been shot just yesterday, as part of some new arthouse project. This goes on to show the prowess of AI and neural networks, and actually highlights both aspects – the bad and the good. While the good would entail the use of neural network algorithms to rescue old and rare films that might not last long, the bad comes with how these tools can be used to effectively doctor footage, in process giving rise to what we today know as deep fakes.

Legend says that the film, which was the first time that most of mankind saw images move on screen, led to the audience running away scared thinking that the train was actually going to run them over. However, after 124 years of continued innovation, we now stand at a point where 8K video recording is commercially and practically feasible, and frame rates go all the way up to 1,000fps in user cameras. Advanced technologies such as the Gigapixel AI and DAIN tools used by Shiryaev here are today acting like a bridge between the past and the future, one which can give film restoration projects a great, new, powerful tool.

Categories
Tech

Google Meena, a New AI Chatbot, Wants to be a Natural at Making Small Talk

Artificial intelligence has grown progressively in complexity, and Meena aims to take it even further. Representative image. (Illustration: Flickr)

Artificial intelligence has grown progressively in complexity, and Meena aims to take it even further. Representative image. (Illustration: Flickr)

Google’s AI Research recently described its latest chatbot, which attempts to be “sensible” in conversation.

  • News18.com
  • Last Updated: February 6, 2020, 6:02 PM IST

With technology prevailing in nearly every single area of life, we often end up making conversations with artificially intelligent algorithms through our day. Think of when you asked Siri to identify a song for you, or Alexa to translate a particularly difficult Hindi word, or even the Google Assistant to read out a message you just received. While all of these features mostly work fine as standalone queries, these voice based conversations do not quite sound natural, or have a flow to them the way they would if you were to ask them to a fellow human being. It is exactly this that Google Meena, the company’s research division’s newest AI chatbot, wants to achieve.

In essence, the way Google Meena differs from other chatbots in the market is by the complexity of conversations that it is capable of. As a human being, our brains are contextually aware of topics, and can intelligently gauge aspects such as a conversation’s tone, to take a dialogue forward. For AI digital assistants, this is still a distant dream. For instance, ask Amazon’s Alexa for how many runs Virat Kohli scored in the previous cricket match, and it is likely to give you the correct response. However, if you follow this up with a second query of “did he play well?”, you’re likely to hit a dead end, with Alexa stating, “I don’t know what that is, but I could look it up for you.” Occasionally, Alexa might also end up playing a song called ‘Well’ from one of the linked streaming services, or in the worst case scenario, lead you to an Amazon listing of a book titled ‘Play Well’.

Such conversations are evidently not very natural. Giving a rather detailed, technical explanation of Meena in Google’s AI Research blog, Daniel Adiwardana and Thang Luong, senior research engineer and scientist (respectively), write, “Meena (is) a 2.6 billion parameter end-to-end trained neural conversational model. We show that Meena can conduct conversations that are more sensible and specific than existing state-of-the-art chatbots. Such improvements are reflected through a new human evaluation metric that we propose for open-domain chatbots, called Sensibleness and Specificity Average (SSA), which captures basic, but important attributes for human conversation. Remarkably, we demonstrate that perplexity, an automatic metric that is readily available to any neural conversational models, highly correlates with SSA.”

In simpler language, Meena has been built on a larger pool of data, computed through more complex networks and ways of structuring sentences. While the full technical explanation, figures and parameters can be found in Google’s blog here, what’s more interesting to note is the scope of usage of such AI algorithms. As a result of its ability to resolve complications, Meena is the first draft of an AI engine that can host a full conversation with you — even what we humans call “small talk”. With this, you can ask the assistant “What do you think of Richard Ashcroft?” Instead of giving a “It’s not my thoughts that matter, it’s yours”, an engine like Meena will probably be able to add context, and offer an opinion of its own.

This is also what adds aspects such as personality and depth to an AI engine — abstract concepts that have so far stayed away from the robots. Barring episodes of Black Mirror, AI development is progressing towards more advanced implementation of natural language processing, and more intricate neural networks that strive to emulate the human brain. Samsung’s Neon, the artificial humans, is one such early straw in a haystack that is sparse now, but bound to grow going forward. This, however, raises many questions about the use of such technologies, and raises the need to discuss the massive mental health and security implications that technology of such stature bring to the table. Multiple research papers have noted the impact of natural voice digital assistants can have on people, and while there have been plenty of positive anecdotes as well, careful safeguarding is required before such tools become mainstream.

Adiwardana and Luong have stated that Meena’s achievements were solely aimed at showcasing specificity and sensibleness, and the future research projects would be looking at aspects such as personality and factuality of information. For now, though, Meena looks to be somewhat adept for a first attempt at small talk, although that’s nowhere near your next door neighbour’s levels of intrusion.

Categories
Tech

This City Uses CCTV Cameras And AI to Shame People Wearing Pyjamas in Public

The Chinese city official has since apologised, the way anyone in the Chinese authority is perhaps expected to.

This City Uses CCTV Cameras And AI to Shame People Wearing Pyjamas in Public
It is expected that China will add 400 million more surveillance cameras across its cities by the end of the year, in addition to the 170 million or so that are already in place. (Images: Twitter and Reuters)

This city in China thinks wearing pyjamas outdoors is “uncivilized behavior”. And this is the city of Suzhou in China. The Government officials in Suzhou in the Anhui province released pictures of seven citizens wearing their nightwear outdoors, on a WeChat account run by the local government and categorized it as uncivilized behavior. The pictures were captured by surveillance cameras installed in the city. And it wasn’t just photos. These were accompanied by details such as the person’s name and ID card numbers. Later, a city official apologized for this uncalled-for naming and shaming in public and on online platforms. As it turns out, Suzhou city is taking part in a national “civilised city” competition, and that residents were banned from wearing pyjamas in public which apparently pull down the ranking of the city.

In a tweet by Vicky Xiuzhong Xu, a researcher at the Australian Strategic Policy Institute says, “People publicly shamed in China for going out in their PJs. Their surname, picture, and partial ID number made public. Local govt asked people to submit more “uncivilized” PJ photos – 10 bucks for each picture verified.” The officials had a justification for this—and this is the closest one should come to expecting an apology from the Chinese, if at all . “We wanted to put an end to uncivilised behaviour, but of course we should protect residents’ privacy,” said the officials said in a statement reported by the BBC.

But if you think this is going to stop, think again. The officials say that they will now blur out the person’s face in the images instead.

It is expected that China will add 400 million more surveillance cameras across its cities by the end of the year, in addition to the 170 million or so that are already in place. Many of these cameras are artificially intelligence, tapping into the massive database that the Chinese authorities have in place to identify citizens using facial recognition.

The scientists at Fudan University working with the Changchun Institute of Optics from the Chinese Academy of Sciences have already developed a 500-megapixel surveillance camera that can identify one person in a crowd of thousands.

Get the best of News18 delivered to your inbox – subscribe to News18 Daybreak. Follow News18.com on Twitter, Instagram, Facebook, Telegram, TikTok and on YouTube, and stay in the know with what’s happening in the world around you – in real time.


Categories
Tech

Microsoft is Bringing its AI Tools to Workplaces, But They Aren’t For Everyone Yet

Microsoft is one of the world’s largest technology firms, so naturally, it invests a fair amount of time, energy and resources on artificial intelligence. Beyond the industrial applications, mainstream tools have been getting a steady entry of predictive tools into their services. Take predictive typing for instance, which is now present in everyday applications such as Microsoft’s Outlook and Google’s Gmail. That, however, only scratches the surface.

The technology, as we know it today, has far greater application in business and industries, but its consumer-end applications have been largely keyword-centric, or behind the scenes, to say the least. With its new AI tools as part of the Microsoft 365 suite, the company wants to make them visible as they operate. In other words, they want to make these tools more relevant to your everyday work. But, while their benefits are apparent, they might not be the best fit for just about anyone.

What they do

Microsoft’s AI tools have a wide range of applications. Aneesh Dhawan, director of solution sales at Microsoft India, gave a smooth demonstration of Microsoft’s AI work tools. While some range from segregating pools of data into neatly charted spreadsheets, others include ways to export a printed sheet straight to digital form without needing to type, and live captioning tools that transcribe and keep both audio-visual and textual records through video conferences. Computer vision, machine learning (ML) and natural language processing (NLP) are at the heart of these technologies, and that’s hardly surprising — if you’ve heard of AI, you’ve likely heard of all three of these as well.

In many ways, Microsoft’s AI tools for workplaces are quite good — impressive, even. For instance, if you have a video call scheduled with a new business partner that natively speaks another language, Microsoft’s new live captioning and transcription tool embedded in Teams will live caption the call right in front of you, easing communication barriers. I saw a glimpse of this during the roundtable arranged by Microsoft, and it did not seem to have any problem in picking up the average Indian English accent. Additionally, the live text transcription of conferences are crucial for someone like me, for whom manually transcribing an interview is excruciatingly painful.

Other tools include Cortana-powered automatic meeting scheduling and handsfree email writing, a visual app that can scan a sheet of paper to automatically export spreadsheets on it, and a corresponding data tool that can further segregate the data from garbled spreadsheets to generate graphs, tabulated sub-sections and more. Further NLP and visual data processing tools in Microsoft PowerPoint can transform scribbled notes into digitised forms. Essentially, from what I gathered from the brief demonstration, Microsoft’s new AI tools are robust and accurate enough to replace big data processing employees, and even EAs at work.

Who they are for

Here, however, is the catch. While everything Microsoft showed seem to work quite well, and I don’t have reason to think they won’t — Big Tech firms do make technologies that are to the point, I’m not entirely sure whether they are quite the right fit for everyone. A majority of Indian workplaces and businesses are small or medium-volume, and many offices primarily work with a limited workforce. For such companies, using a free to use service such as Google Drive, or even Microsoft’s own OneDrive, often suffice. In contrast, Microsoft’s present pricing pegs the Microsoft 365 suite at a basic price of Rs 125 per user per month, with the full productivity suite priced at Rs 1,320 per user per month.

Then comes the need for a uniform ecosystem — for all of Microsoft’s tools to work, it is important that everyone in your organisation uses a Microsoft system, and is subscribed to the Microsoft 365 suite. While this is something that a service provider should not be penalised for, it’s important to note that most workplaces that operate on a smaller scale in India mostly stick to WhatsApp or Slack for communication, and the aforementioned services for the rest. The question hence remains is whether these tools are attractive enough to warrant even those working on freeware to switch to paid services.

From the preview roundtable, Microsoft’s AI tools seem great, but tailored for larger organisations churning heavy data and hosting cross-language conferences on an everyday basis. While collaborative and digital workspaces are on the rise, Microsoft’s AI tools alone may not be the unique selling proposition to urge a business to join the company’s ecosystem.

What this means

To sum up, Microsoft may have a winner in its hands if one considers the rising need to digitise years of paperwork. Questions of privacy will be answered at a later stage, but for now, Microsoft vouches to its cybersecurity credentials. It also states that any data snippets collected from its services are done anonymously, and kept within the confines of a “select” few nations. It also states that any data collected and operated in India, stays within Microsoft’s India data servers.

However, for the time being, Microsoft’s AI tools and services in the Microsoft 365 suite are better suited, by design, to large businesses. They are great at what they do, but they may not be the ones that definitively alter the definition of human effort at the average everyday workplace.

Categories
Tech

Microsoft Disagrees With EU And Sundar Pichai on Temporary Ban of Face Recognition Tech

The EU’s proposal for a temporary ban on facial-recognition technology won backing from Alphabet Chief Executive Sundar Pichai but got a cool response from Microsoft President Brad Smith. While Pichai cited the possibility that the technology could be used for nefarious purposes as a reason for a moratorium, Smith said a ban was akin to using a meat cleaver instead of a scalpel to solve potential problems.

“I think it is important that governments and regulations tackle it sooner rather than later and give a framework for it,” Pichai told a conference in Brussels organised by think-tank Bruegel. “It can be immediate but maybe there’s a waiting period before we really think about how it’s being used,” he said. “It’s up to governments to chart the course” for the use of such technology. Smith, who is also Microsoft’s chief legal officer, however cited the benefits of facial recognition technology in some instances such as NGOs using it to find missing children. “I’m really reluctant to say let’s stop people from using technology in a way that will reunite families when it can help them do it,” Smith said. “The second thing I would say is you don’t ban it if you actually believe there is a reasonable alternative that will enable us to, say, address this problem with a scalpel instead of a meat cleaver,” he said.

Smith said it was important to first identify problems and then craft rules to ensure that the technology would not be used for mass surveillance. “There is only one way at the end of the day to make technology better and that is to use it,” he said. The European Commission is taking a tougher line on artificial intelligence (AI) than the United States that would strengthen existing regulations on privacy and data rights, according to a proposal paper seen by Reuters. Part of this includes a moratorium of up to five years on using facial recognition technology in public areas, to give the EU time to work out how to prevent abuses, the paper said.

Pichai urged regulators to take a “proportionate approach” when drafting rules, days before the Commission is due to publish proposals on the issue. Regulators are grappling with ways to govern AI, encouraging innovation while trying to curb potential misuse, as companies and law enforcement agencies increasingly adopt the technology. There was no question AI needs to be regulated, Pichai said, but rulemakers should tread carefully. “Sensible regulation must also take a proportionate approach, balancing potential harms with social opportunities. This is especially true in areas that are high risk and high value,” he said.

Regulators should tailor rules according to different sectors, Pichai said, citing medical devices and self-driving cars as examples that require different rules. He said governments should align their rules and agree on core values. Earlier this month, the U.S. government published regulatory guidelines on AI aimed at limiting authorities’ overreach, and urged Europe to avoid an aggressive approach. Pichai said it was important to be clear-eyed about what could go wrong with AI, and while it promised huge benefits there were real concerns about potential negative consequences.

One area of concern is so-called “deepfakes” – video or audio clips that have been manipulated using AI. Pichai said Google had released open datasets to help the research community build better tools to detect such fakes. The world’s most popular internet search engine said last month that Google Cloud was not offering general-purpose facial-recognition application programming interfaces (APIs) while it establishes policy and technical safeguards.

Categories
Tech

Alphabet CEO Sundar Pichai Insists on Regulating AI, But is Google Doing Enough?

Sundar Pichai, the chief executive of Alphabet and Google, has reiterated his stance on the need for regulation of artificial intelligence technologies, in an op-ed published in The Financial Times. Stating the importance of regulating a technology as widespread and powerful as AI, Pichai brought up Google’s AI principles published in 2018, and stated that the technology makers now need to work with policymakers to realise the full potential of technologies such as AI, and develop them to their true abilities.

In recent times, the use of AI has, as with most powerful innovation, seen considerable usage in non-ideal situations. It is this that warrants cross-sector and cross-industry collaboration, writes Pichai. He states, “(The) history is full of examples of how technology’s virtues aren’t guaranteed. These lessons teach us that we need to be clear-eyed about what could go wrong. There are real concerns about the potential negative consequences of AI, from deepfakes to repressive uses of facial recognition. While there is already some work being done to address these concerns, there will inevitably be more challenges ahead that no one company or industry can solve alone.”

Pichai described the use of Google’s AI tools in critical sectors, taking examples from healthcare (usage of AI by doctors in early detection of breast cancer symptoms), climate change (use of AI to pre-detect weather anomalies) and travel (use of AI to reduce flight delays). He further touched upon how Google uses its new AI tools to see their compliance with their own AI principles, stating, “Principles that remain on paper are meaningless. So we’ve also developed tools to put them into action, such as testing AI decisions for fairness and conducting independent human-rights assessments of new products. We have gone even further and made these tools and related open-source code widely available, which will empower others to use AI for good. We believe that any company developing new AI tools should also adopt guiding principles and rigorous review processes.”

However, many have, over time, criticised Google’s principles for not being transparent enough, and even for their stance on accountability of their own technologies. While Pichai speaks in the op-ed about testing his company’s technologies, many have stated that Google needs to be more accountable in how their technology works, and just self-regulatory principles do not make the cut. For instance, Google’s recent work on using AI to detect deepfakes on the web only builds a database that can be tallied to detect the problem, but does not provide preventive measures to curb its spread, beyond the scope of this database. While arguments regarding infrastructural limitations will remain, a company of Google’s magnitude and technological prowess have time and again faced the criticism of not doing enough on the final third.

It is this that brings light to Pichai’s push for government regulations, and building uniform international regulations. Official regulations, as seen often, typically develops at a far slower pace in comparison to technologies such as AI, which has progressed at a staggering speed. As a result, it is questionable as to how far can the idea of uniform government regulations across the world in terms of ethical and legal implementation of AI across sectors can be framed. Rather, while framing such regulations is technically possible, it is too difficult to establish such regulations within a short span of time, bringing the onus of self-regulation and transparency back to tech giants such as Google.

“AI has the potential to improve billions of lives, and the biggest risk may be failing to do so,” says Pichai in the FT op-ed. While pretty much everyone will agree to it, what remains to be answered is how much of that would depend on the responsibility that lies of Google’s (and others’) own shoulders.


Categories
Tech

Apple Acquires Xnor.ai, an Edge-based AI Startup, for $200 Million

Apple’s acquisition of Xnor.ai is believed to be a routine startup acquisition in a bid to incorporate new technologies into its own service.

I

Apple Acquires Xnor.ai, an Edge-based AI Startup, for $200 Million
Image for Representation

Apple has acquired Xnor.ai, a Seattle based startup specialised in low-power, edge-based artificial intelligence tools for around $200 million. “Apple buys smaller technology companies from time to time and we generally do not discuss our purpose or plans.” said Apple in a statement given to GeekWire. Xnor.ai’s work is expected to be be incorporated into future iPhones, improving Siri and other AI and machine learning-based tasks. Recently, Apple has acquired another Seattle-based machine learning startup Turi for almost same amount. The move is set to increase Apple’s presence in the Seattle region where the tech giant has been building an engineering outpost

Turi offers tools that are meant to let developers easily scale machine learning applications. The Cupertino-based company has recently bought some machine learning and AI startups like VocalIQ and Perceptio and facial recognition startup Emotient, among others. Apple has recently been making a push into artificial intelligence through Siri personal assistant and related technologies.

Categories
Tech

Microsoft Math Solver App Uses AI to Solve Those Concepts You Can't Make Sense of

The app will use artificial intelligence to recognize and provide instant as well as step-by-step solutions for a wide range of math concepts.

Microsoft Math Solver App Uses AI to Solve Those Concepts You Can't Make Sense of
Microsoft Math Solver App

(Image: Microsoft)
(Image altered by News18)

Microsoft has come out with a new app that could be a lifesaver, especially if you are a student. The Math Solver app will be able to help solve every possible math problem, be it simple addition or differentiation. The application will work on every Android or iOS device and can be downloaded from the Google Play Store or the Apple App Store. According to its description, the app supports 22 languages including 12 Indian languages such as Assamese, Bengali, Gujarati, Hindi, Kannada, Konkani, Marathi, Malayalam, Oriya, Punjabi, Tamil, and Telugu. International languages such as German, Spanish, Simplified Chinese, and Russian will also be compatible with the app.

The app will work on the basis of artificial intelligence (AI) and recognize math problems and then provide solutions for a wide range of concepts. In order to use it, you can simply add your problem by drawing on the screens. Or, you can type the problem. Apart from these, you can also upload a scanned picture of the problem written on a book or notebook. This simple procedure will upload your problem and the app will work on solving it.

Using its AI algorithms, it will recognise the text and display the problem on the screen. Following this, the Microsoft app will provide you with “instant solutions”. It will highlight a step-by-step guide for solving the problems to teach you how to solve it rather than depending on the app. Even, graphs will be used to make you understand math problems. The app also comes with an in-built dark mode.