Daily Archives: September 1, 2021

Steve Dent is an associate editor at Engadget.More posts by this contributor

The voices on Amazons Alexa, Google Assistant and other AI assistants are far ahead of old-school GPS devices, but they still lack the rhythms, intonation and other qualities that make speech sound, well, human. NVIDIA has unveiled new research and tools that can capture those natural speech qualities by letting you train the AI system with your own voice, the company announced at the Interspeech 2021 conference.

To improve its AI voice synthesis, NVIDIAs text-to-speech research team developed a model called RAD-TTS, a winning entry at an NAB broadcast convention competition to develop the most realistic avatar. The system allows an individual to train a text-to-speech model with their own voice, including the pacing, tonality, timbre and more.

Another RAD-TTS feature is voice conversion, which lets a user deliver one speakers words using another persons voice. That interface gives fine, frame-level control over a synthesized voices pitch, duration and energy.

Using this technology, NVIDIAs researchers created more conversational-sounding voice narration for its own I Am AI video series using synthesized rather than human voices. The aim was to get the narration to match the tone and style of the videos, something that hasnt been done well in many AI narrated videos to date. The results are still a bit robotic, but better than any AI narration Ive ever heard.

With this interface, our video producer could record himself reading the video script, and then use the AI model to convert his speech into the female narrators voice. Using this baseline narration, the producer could then direct the AI like a voice actor tweaking the synthesized speech to emphasize specific words, and modifying the pacing of the narration to better express the videos tone, NVIDIA wrote.

NVIDIA is distributing some of this research optimized to run efficiently on NVIDIA GPUs, of course to anyone who wants to try it via open source through the NVIDIA NeMo Python toolkit for GPU-accelerated conversational AI, available on the companys NGC hub of containers and other software.

Several of the models are trained with tens of thousands of hours of audio data on NVIDIA DGX systems. Developers can fine tune any model for their use cases, speeding up training using mixed-precision computing on NVIDIA Tensor Core GPUs, the company wrote.

Editors note: This post originally appeared on Engadget.

Read more:

NVIDIAs latest tech makes AI voices more expressive and realistic - TechCrunch

The synergies of AI and IoT are now making serious headway in the automotive, industrial and medical sectors, according to Omdias Josh Builta.

When it comes to the digital transformation, the smart money is likely to follow IoT applications that integrate with artificial intelligence (AI), according to Josh Builta, research director, Internet of Things, Omdia.

Speaking to IoTWTs parent company Informa Tech at DesignCon 2021, Builta predicted AI would emerge as the practical means to make sense of output from IoT endpoints, forecast to reach 75 billion by the end of the decade.

Theres a virtuous feedback loop, too, Builta added, as AI enhances the conclusions of connected devices but it also refines its judgment based on the data that IoT extracts.

Its hard to imagine any human being able to make sense of that data. And its not only the amount of data, but also whether [enterprises] are able to trust the data at the end, Builta told Chuck Martin, editorial director of AI and IoT at Informa Tech.

While Martin alluded to the reality that credibility issues also arise from AI-driven analysis, Builta posited that quality would likely improve as automation models ingest more data.

With both AI and IoT being key to the digital transformation, Omdia has seen many enterprises accelerating deployments to make the most of connected technology during the COVID-19 pandemic.

Builta cited connected vehicles, industrial and health care as areas where AIoT was already prevalent. A connected car is essentially an IP address, but when you add AI into that you can allow for these vehicles to operate without a human interface. These are the types of examples that were now starting to see.

Here is the original post:

AI and IoT to Ignite Digital Transformation - IoT World Today

Weve had visionary investors onstage before, and weve had science fiction authors onstage but never at the same time, let alone a pair who collaborated on a unique book of stories and essays that make an optimistic prediction of our AI-infused future. Sinovation founder Kai-Fu Lee and author of Waste Tide and others Chen Qiufan will join us at Disrupt (September 21-23) for a discussion of the fiction and fact of todays hottest technology.

Lee, born in Taiwan, attended CMU and obtained a PhD in computer science, working initially on speech recognition before working for Apple, SGI and Microsoft, then establishing Google China as its president. His research and investment company, Sinovation (originally Innovation Works) has been his focus since its founding in 2009, and he has grown to become a leading mind and influential figure in AI.

When we last spoke with Lee, at Disrupt SF 2018, he emphasized that China was catching up to the U.S. on AI research, and had surpassed it in some ways. And certainly his own investments have contributed to that. Since then, as someone who thinks frequently about what the future holds, he has found a kindred spirit in Chen Qiufan.

Qiufan is a Chinese author whose 2013 novel Waste Tide propelled him to literary fame, though like many authors, that wasnt enough to make him quit his day job until a few years later (Wired only just ran a profile on him). But by that time he had attracted the attention of Lee, who proposed a novel project: a collaborative book where the two would put their heads together to create a fictitious future informed by fact and realistic extrapolation.

The result is AI 2041: 10 stories by Qiufan set in the titular year, all over the world, with people from all walks of life encountering AI in the many ways that the authors speculate it may come to shape society over the next two decades. Each is followed by an explanatory essay by Lee that goes into the technical aspects and why they might lead to that future.

Ill be posting a full review of the book ahead of the event, but I can certainly say that its unlike any collection Ive read before. Each story is independent but takes place in something like a shared world, and each illustrates a potential application, conflict or change in thinking that AI could lead to. And, importantly, the AI is recognizable as descended directly from existing technologies.

For instance, one story concerns a talented deepfake creator working out of Lagos, one who knows the ins and outs of generative adversarial networks, image inspection, media pathways and so on. Hes tasked with creating a video of a long-dead celebrity that fools not just people watching it but the hosting services automated scanners, the governments facial recognition algorithms and all the rest but he begins to suspect theres an unsavory motive behind it all (I wont spoil the rest).

What follows the story is Lees essay on GANs, facial recognition and deepfakes that explains the concepts in an understandable but not patronizing way, then explores the risks and benefits in a non-narrative way. It helps ground the stories as real possibilities, not just imagined situations.

With both Qiufan and Lee onstage (virtually this time), the discussion of the book and the issues it brings up should be a lively one not least because it will be moderated by yours truly. But to catch this session, youll need to grab a pass to attend Disrupt happening September 21-23. Get yours today for less than $100 for a limited time!

View post:

Kai-Fu Lee and Chen Qiufan will share their vision of our AI-powered future at Disrupt - TechCrunch

(EtiAmmos/Shutterstock)

While there is an increase in the number of investigations companies are conducting in the areas of employee conduct, regulatory compliance, security, and privacy, companies by and large are not making full use of the advanced analytics and AI tools at their disposal, according to a new report by OpenText.

OpenText commissioned Compliance Week to conduct the study, which is based on a survey of 200 compliance, legal, and internal audit professionals. The study found that investigations across employee conduct, regulatory compliance, security, and privacy have increased between 14% and 32% over the past year.

But the investigators are running into obstacles along the way, according to the survey, which found 42% faced time restraints, 39% experienced difficulty collecting data from remote and off-network locations, and another 39% ran into trouble collecting data from new sources of electronically stored information.

Despite the increase in the number of investigations expected to take place over the next two years, budgets are largely expected to be flat over that time, the survey says. Thats a concern, especially considering the lack of automation being applied to the field.

(Source: Compliance Week)

In fact, OpenText reports that 76% of survey responders will use a manual approach to gathering and analyzing data. Just over half (56%) will conduct keyword search and linear batching review, the company says. The report found that 31% will use technology-assisted review or machine learning, and 30% will use advanced analytics, to help them with their work.

The results of the study show that, despite a perceived need for improved efficiencies and better outcomes, companies just arent using analytics and AI to investigate the large amount of electronically stored information.

Data analytics, automation, and machine learning are necessary tools in supporting investigations, Lou Blatt, OpenTexts senior vice president and chief marketing officer says in a press release. Vast increases in information, changing data privacy and compliance requirements, and growing cybersecurity risks are all contributing to the need for a faster approach to managing and conducting investigations that results in better outcomes.

OpenText recommends that companies not only adopt AI and analytics, but also adopt a strike team approach to more efficiently tackling compliance and audit investigations. It further recommends that companies pair these strike teams with a group of existing business leaders in HR, compliance, legal, audit, risk, security, and IT.

You can access the Compliance Week report, which is the subject of a webinar on September 23, at this link.

Related Items:

AI Makes Inroads into Enterprise Software

The Sunny Side of Privacy Laws and Compliance Mandates

5 Ways Analytics Drives Cyber Monday Sales

More:

Analytics and AI Not Being Fully Utilized for Audits and Compliance Investigations: Report - Datanami

MARKHAM, ON, Aug. 31, 2021 /CNW/ - Canadian Artificial Intelligence (AI) company EAIGLE Inc., specializing in COVID-19 solution technology, has launched a new proof of vaccination platform to provide its existing five (5) million monthly users with an additional layer of health and safety protection. The new feature, which integrates with its portfolio of wellness screening solutions, will be available to clients by September 30th, 2021.

EAIGLE Visitor Management & Wellness Screening With Digital Vaccine Pass Platform (CNW Group/EAIGLE)

"We recognize the growing need to protect workplaces and public spaces via a flexible and automated solution that is reliable, easy to implement, and scalable." said, Amir Hoss, EAIGLE's CEO.

"This is why we have designed a platform that's in line with current market expectations but can easily evolve to meet the needs of a dynamic landscape," Hoss added.

ABOUT EAIGLE'S DIGITAL VACCINE PASS PLATFORM

Governments are making concerted efforts to reopen cities and businesses while new threats continue to emerge. As new variants of COVID-19 are detected, leveraging technology will be vital to ensure a smooth return to normalcy in the workplace.

EAIGLE's Digital Vaccine Pass is a proof of vaccination platform that enables governments and organizations to verify vaccination status at scale. It empowers users to upload their proof of vaccination online or scan it on-site at EAIGLE's wellness stations through a touchless and automated process.

To learn more about EAIGLE's Digital Vaccine Pass, visit: http://www.eaigle.com/digital-vaccine-pass

PRODUCT INNOVATION APPROACH

EAIGLE believes that collaboration with its clients is imperative to new product innovation. Working with our partners has given us unique insights into the needs and challenges of organizations across industries as they design new processes to safeguard their workplaces. Our first wave of loyal clients and innovation partners has helped shape the product development process for EAIGLE's visitor management and wellness screening solutions. We used the same approach to design EAIGLE's Digital Vaccine Pass Platform. By engaging with our clients, we've been able to validate many product assumptions and ensure a market-ready platform at launch.

Story continues

About EAIGLE

EAIGLE is a leader in artificial intelligence and computer vision technology, designing next-generation solutions to address the most complex challenges faced by a variety of industries. From streamlining operational processes to creating the technology layer for smart buildings, EAIGLE's technology helps public and private sector organizations make smarter decisions that enable them to future-proof their operations.

Since the start of the pandemic, EAIGLE has been working with organizations to mitigate disruptions at work and in public spaces in the fight against COVID-19 in an effort to maintain business continuity. The company's deep expertise in AI technology is underpinned by a commitment to high-quality software development through constant innovation and investment in R&D, automation, training, testing, and support. Today, EAIGLE is one of Canada's fastest-growing AI companies, with a customer footprint that spans across North America.

SOURCE EAIGLE

Cision

View original content to download multimedia: http://www.newswire.ca/en/releases/archive/August2021/31/c1594.html

Link:

Canadian AI technology company EAIGLE launches a new proof of vaccination platform to provide organizations with enhanced health and safety protection...

"LXT is committed to providing the customized solutions that have made our clients successful." - Mohammad Omar, CEO

According to IDC, the global AI market is expected to reach more than $550 billion with a five-year compound annual growth rate (CAGR) of 17.5% by 2024. Demand has never been greater, and to scale globally and take advantage of this opportunity, organizations need to train AI technology on data that captures the unique cultural and linguistic nuances of every region.

"We are pleased to welcome Phil, Asser and Jodie to our executive team as we pursue our vision to provide the data that powers a much more intelligent and automated world," said Mohammad Omar, LXT founder and CEO. "All three have enterprise leadership experience at companies such as Appen and Shaw Communications that will help to fuel our next phase of growth. While many providers are focused on one-size-fits all solutions, LXT is committed to providing the customized solutions that have made our clients successful and these will remain core to our business."

Phil Hall brings more than 20 years' experience in working with the world's top technology companies and has extensive background in speech and linguistics, AI and machine learning data. Asser ElShanawany has a wealth of experience in leading public organizations, transforming and scaling start-ups to large technology companies, particularly in the telecommunications sector. Jodie Ruby comes to LXT with over 20 years of B2B technology marketing experience, including building and leading marketing teams from the ground up for high-growth organizations.

"LXT is a trusted partner to some of the world's largest technology companies - enabling them to deliver cutting-edge AI applications - and has achieved impressive growth based on its ongoing relationships with these organizations," said Phil Hall, LXT's new Chief Growth Officer. "I am thrilled to join a team with such a strong reputation for agile, reliable and cost-effective delivery of high-quality AI data, and look forward to helping LXT capitalize on its huge growth potential."

For enterprises developing AI-based solutions at scale, the ability to collect and annotate the data needed to train these solutions is limited and slows the pace of innovation. Data annotation requires countless hours of human intelligence and translation while data hungry businesses need reliable data collection and annotation partners with global coverage.

"I am excited to join LXT to help establish the company as a leading player in AI training data in partnership with global top 10 technology organizations," commented LXT's new Chief Financial Officer Asser ElShanawany. "Since we are already profitable, debt free and cash flow positive, our immediate focus is on scale, recapitalization, and many other promising financial milestones on the horizon."

LXT provides data generation, collection and annotation services in more than 200 languages through a secure technology platform that facilitates human insight to improve accuracy and streamlines workflow to reduce costs and optimize turnaround times.

"With the acceleration of AI investment across a wide range of use cases and industries, the demand for high-quality, human-annotated data is stronger than ever," said Jodie Ruby, LXT's new Vice President of Marketing. "LXT has proven itself as a trusted partner to leading global enterprises, and I am excited to join this talented team to help build on the momentum it has already created."

LXT was chosen as the exclusive data collection and annotation partner for the SUPERB program, working alongside leading researchers from National Taiwan University, MIT, Carnegie Mellon University, Johns Hopkins University, and Facebook AI. Its stated goal is tofuel research in representation learning and general speech processing. Learn more here.

"SUPERB is a unique effort to create a benchmark for models across a wide variety of tasks and will benefit the broader speech industry by enabling the detection of emotion, intent, content and other semantic information," commented Hung-yi Lee, an associate professor of the Department of Computer Science & Information Engineering at National Taiwan University. "High-quality data is key to the success of this effort, and LXT was chosen as the exclusive partner based on its flexibility, reliability, and collaborative culture."

LXT data services are delivered through its crowdsourced workforce in more than 80 countries around the world, its own secure facilities or onsite client deployment. To meet the most stringent security requirements, LXT facilities are ISO 27001 certified and PCI DSS compliant, and offer supervised annotation to safeguard customer data.

About LXTLXT is an emerging leader in AI training data to power intelligent technology for global organizations, including the largest technology companies in the world. In partnership with an international network of contributors, LXT collects and annotates data across multiple modalities with the speed, scale and agility required by the enterprise. Our global expertise spans 80countries and over 200 languages. Founded in 2014, LXT is headquartered in Toronto, Canada with presence in the United States, Australia, India, Turkey and Egypt. The company serves customers in North America, Europe, Asia Pacific and the Middle East. Learn more at lxt.ai.

SOURCE LXT

http://www.lxt.ai

Visit link:

LXT to Capitalize on Massive Opportunity in AI Training Data with Agile and Customized Solutions, Expanded Management Team and Exclusive Partnership -...

The EUs proposed AI legislation published in April, sparked debate on the true impact that new AI rules would have on businesses. Overall, it seems to be that the legislation has the potential to benefit society as a whole, but this could ultimately hinder companies and how they use AI in the long term.

However, OReillys recent AI in the Enterprise research discovered that, while more businesses are continuing to use AI or are considering implementing it in the near future, only 52 percent of these companies are checking for issues of fairness or bias within their AI systems.

One of the major roadblocks to AIs advancement has been a lack of trust in the technology. This is especially true in the public sector, where AI-assisted choices may have a significant influence on peoples lives. The EUs AI legislation aims to correct this by assisting organizations in navigating ethical AI usage. This will help to establish trust over time, allowing businesses to ultimately realize AIs full potential. Businesses must now change how they use and implement AI to ensure that they always fall on the right side of the line.

The AI-train has rapidly been gaining momentum in recent years, both in terms of business usage and results. Were now seeing the technology being used for cancer detection, climate change analysis, the control of traffic and marketing for businesses. Globally, a quarter (26 percent) of businesses have reached the mature stage of AI usage. This means that they have revenue-yielding AI products in production. In the UK, this figure is even higher, with 36 percent classifying their AI usage as mature.

Looking at the industry breakdown, retail came out on top, with 40 percent claiming that their usage of AI was mature. This was closely followed by financial services (38 percent) and telecommunications (37 percent). Comparatively, education (10 percent) and government (16 percent) were the least mature in their usage of AI.

The stats suggest that, while AI adoption in the private sector is snowballing, the public sector is struggling to keep up. The question is: why?

There is likely more than one factor as to why the public sector is struggling in its uptake of AI. Budgetary concerns could certainly be a key issue, but perhaps not enough to account for such a large difference between the public and private sector. The other glaring issue is public trust.

The general public already had their guard up against the use of AI in the public sector. Their worst fears were then proven correct in 2020 when A-Level and GCSE grades were predicted using an AI algorithm that faced accusations of bias. This led to the results being scrapped and replaced by predicted grades given by teachers. Its examples like these which damage public trust in AI.

In terms of checking AI models for bias, the UK is ahead of the global standard. Across the globe, just 52 percent of companies are checking their algorithms for bias. Meanwhile, in the UK, this figure rises to 56 percent. However, when it comes to decisions that impact peoples lives and their futures, a little better than half isnt enough. This counts for both the public and the private sector. Private sector companies, such as banks, also have the power to make decisions that can impact peoples lives.

The EUs AI legislation, which focuses heavily on AI ethics, should force companies to confront these shortcomings and be the starting point for organizations to build public trust and, in time, release the handbrake which is holding AI back. A more educated approach to AI will be key to achieving this.

Its clear that not enough businesses are checking for bias in their AI models. However, research suggests that this isnt necessarily negligence but, instead, a lack of training and skills. Globally, the biggest bottlenecks to AI adoption are a lack of skilled people (19 percent) and data quality (18 percent). In the UK, a quarter (25 percent) labeled a lack of data/data quality as a major hindrance and 14 percent said the same about skills within the organization.

This skills gap is already having a huge impact on the adoption of AI and, with the introduction of the EUs AI legislation, will have an even greater impact if businesses do not act soon. Half of UK businesses admitted that only about 50 percent of their AI projects are actually completed. Meanwhile, as weve seen, those that are completed run a risk of being biased. Moving forward, neither of these options will be profitable for companies.

To close this skills gap, businesses must ensure that they are providing adequate training for their AI-handling employees. This means equipping them with the necessary knowledge to develop and train an algorithm that is highly functional and ethical. Feeding the algorithm with high-quality and unbiased data is the first step, but employees must also be trained to consistently check the algorithm for bias or inconsistencies and make the necessary changes.

With the introduction of the new AI laws, some employees may be nervous to make a mistake. Businesses can take this fear away by empowering their employees to learn in the flow of work. This means allowing them to ask questions and receive quick answers, based on the most up-to-date guidance, which they can apply to their work. The learning platforms to enable this exist, and its now time for employers to start leaning on them. Or they could be one of the first organizations to feel the sting of the new AI legislation.

Businesses and organizations may be tempted to interpret the new AI regulation as a restriction on their technological ambitions. Instead, it should be viewed as advice that will assist them in making the most out of AI. Companies can roll out AI initiatives without fear of public backlash if they stay inside the confines of the new legislation. This will then enable them to test new AI technologies with greater confidence in the long term. However, to build this trust, businesses need to continually keep getting it right when it comes to AI. This means no more instances of AI bias or technologies which push the boundaries of privacy. Regular education and training is the only way to achieve this level of continued excellence.

Rachel Roumeliotis, Vice President of Data and AI, OReilly

Read the rest here:

A spotlight on the EUs AI legislation Realizing the full potential of AI - ITProPortal

Humans are active listeners; we create meaning where there is none, or none intended. It is not that the octopuss utterances make sense, but rather that the islander can make sense of them, Bender says.

For all their sophistication, todays AIs are intelligent in the same way a calculator might be said to be intelligent: they are both machines designed to convert input into output in ways that humanswho have mindschoose to interpret as meaningful. While neural networks may be loosely modeled on brains, the very best of them are vastly less complex than a mouses brain.

And yet, we know that brains can produce what we understand to be consciousness. If we can eventually figure out how brains do it, and reproduce that mechanism in an artificial device, then surely a conscious machine might be possible?

When I was trying to imagine Roberts world in the opening to this essay, I found myself drawn to the question of what consciousness means to me. My conception of a conscious machine was undeniablyperhaps unavoidablyhuman-like. It is the only form of consciousness I can imagine, as it is the only one I have experienced. But is that really what it would be like to be a conscious AI?

Its probably hubristic to think so. The project of building intelligent machines is biased toward human intelligence. But the animal world is filled with a vast range of possible alternatives, from birds to bees to cephalopods.

A few hundred years ago the accepted view, pushed by Ren Descartes, was that only humans were conscious. Animals, lacking souls, were seen as mindless robots. Few think that today: if we are conscious, then there is little reason not to believe that mammals, with their similar brains, are conscious too. And why draw the line around mammals? Birds appear to reflect when they solve puzzles. Most animals, even invertebrates like shrimp and lobsters, show signs of feeling pain, which would suggest they have some degree of subjective consciousness.

But how can we truly picture what that must feel like? As the philosopher Thomas Nagel noted, it must be like something to be a bat, but what that is we cannot even imaginebecause we cannot imagine what it would be like to observe the world through a kind of sonar. We can imagine what it might be like for us to do this (perhaps by closing our eyes and picturing a sort of echolocation point cloud of our surroundings), but thats still not what it must be like for a bat, with its bat mind.

Excerpt from:

What would it be like to be a conscious AI? We might never know. - MIT Technology Review

SUNNYVALE, Calif.--(BUSINESS WIRE)--FogHorn, a leading developer of Edge AI software for industrial and commercial Internet of Things (IoT) solutions, today announced its ranking as the Overall Leader, Top Innovator and Top Implementer by ABI Researchs competitive vendor assessment on IoT Edge Analytics: Hardware-Agnostic SaaS and PaaS.

ABI Research assessed vendors based on their deployment of advanced edge analytics and artificial intelligence (AI) technologies to customers in various industries, go-to-market strategies, scalability and efficiency. The competitive ranking offers an unbiased assessment of edge-cloud software-as-a-service (SaaS) and platform-as-a-service (PaaS) technologies enabling the Internet of Things (IoT) for enterprises, covering vendors that are providing hardware agnostic machine learning (ML) and AI.

Leveraging edge intelligence enables enterprises to achieve operational efficiency, reduce costs and enhance workplace and asset monitoring, said Chris Penrose, Chief Operating Officer at FogHorn. Were honored to be recognized by ABI Research for enabling our customers to reach these goals with our Lightning Edge AI Platform and Solutions. This competitive assessment showcases the value were driving for our customers and highlighting a variety of edge AI use cases that ultimately enhance their decision-making and ROI with data-driven insights.

FogHorn established itself as the leader due to its performance in the industrial vertical and wide range of clients and strategic partnerships. Additionally, ABI Research noted FogHorns ability to serve multiple IoT use cases and sophisticated capabilities for predictive analytics and ML as a key consideration of its evaluation. As highlighted by ABI Research, FogHorn received higher implementation scores because of its influence and adoption rate of video analytics for the IoT domain.

FogHorn was also ranked as a vendor successfully monetizing market opportunities resulting from the COVID-19 pandemic. In June 2020, FogHorn announced its Health and Safety monitoring solution, which enables enterprises to address employee wellbeing and help prevent exposure to COVID-19 and monitor workplace safety through personal protective equipment detection and hazard monitoring. This solution, delivered as a ready-to-use package that utilized ML combined with video analytics, diversified FogHorns product portfolio compared to other edge AI vendors by ABI Research.

FogHorns Lightning Edge AI Platform was the first edge-native AI solution built for secure, on-site intelligence. Its edge processing capabilities are ideal for low-latency use cases enabling real-time data processing harnessing ML and AI capabilities within a minimal compute footprint. In addition to its recognition as an Overall Leader, FogHorn earned a top mark for predictive and ML modeling, as well as measurement against ABI Researchs unique innovation criteria.

Download a copy of ABI Researchs competitive assessment ranking on IoT Edge Analytics: Hardware-Agnostic SaaS/PaaS from the FogHorn website here.

About FogHorn

FogHorn is a leading developer of edge AI software for industrial and commercial IoT application solutions. FogHorns software platform brings the power of advanced analytics and machine learning to the on-premises edge environment enabling a new class of applications for advanced monitoring and diagnostics, machine performance optimization, proactive maintenance, and operational intelligence use cases. FogHorns technology is ideally suited for OEMs, systems integrators and end customers in manufacturing, power and water, oil and gas, renewable energy, mining, transportation, healthcare, retail, as well as smart grid, smart city, smart building, and connected vehicle applications.

See the original post here:

FogHorn and Lightning Edge AI Platform Recognized as Overall Leader by ABI Research - Business Wire

31 August 2021

Sheppard Mullin Richter & Hampton

To print this article, all you need is to be registered or login on Mondaq.com.

In recent years, AI patent activity has exponentially increased.The figure below shows the volume of public AI patent applicationscategorized by AI component in the U.S. from 1990-2018. The eightAI components in FIG. 1 are defined inan article published in 2020by theUSPTO. Most of the AI components have experienced explosive growthin the past decade, especially in the areas of planning/control andknowledge processing (e.g., using big data in automatedsystems).

Figure 1. AI patent activities byyear

AI technology is complex and includes different parts acrossdifferent fields. Inventors and patent attorneys often face thechallenge of effectively protecting new AI technology development.The rule of thumb is to focus the patent protection on what theinventors improve over the conventional technology. However,inventors often need to improve various aspects of an existing AIsystem to make it fit and work for their applications. In thefollowing sections, we will discuss an illustrative list of subjectareas that may offer patentable AI inventions.

The training phase of an AI system includes most of the excitingtechnical aspects of machine learning algorithms exploring thelatent patterns embedded in the training data. A typical trainingprocess includes preparing training data, transforming the trainingdata to facilitate the training process, feeding the training datato a machine learning model, fitting (training) the machinelearning model based on the training data, testing the trainedmachine learning model, and so on. Different AI models or machinelearning models may have different training processes, such assupervised training based on labeled training data, unsupervisedtraining that infers a function to describe a hidden structure fromunlabeled training data, semi-supervised training based onpartially-labeled training data, reinforcement learning (RL), etc.Common areas in the training phase that may yieldpatent-protectable ideas include:

The application phase of an AI system includes applying thetrained models to make predictions, inferences, classifications,etc. This phase generally covers the real application of the AIsystem. It can provide easier infringement detectability and thusvaluable patent protection for the AI system. In this digital era,AI systems can be applied to almost every aspect of our life. Forexample, an AI patent can claim or describe how the AI system helpsthe user to make better decisions or perform previously impossibletasks. These applications may be deemed as practical applicationsthat are powerful in overcoming potential "abstract idea"rejections during the prosecution of the AI patent.

On the other hand, simply claiming an AI system as a magicalblack box that generates accurate predictions based on input datawill likely trigger rejections during prosecution, such aspatentable subject matter rejections (e.g., a simple application ofthe black box may be categorized as human activities). There arevarious ways to reduce the chances of getting such rejections. Forexample, adding a brief description of the training process or themachine learning model structure helps overcome U.S.C. 101rejections.

Another flavor of AI patents is related to accelerators,hardware pieces with built-in software logic accelerating trainingand/or inferencing process. These AI patents may be claimed fromeither a software perspective or hardware perspective. Someexamples include specially designed hardware to improve trainingefficiency by working with GPU/TPU/NPU/xPU (e.g., by reducing datamigrations among different components/units), memory layout changesto improve the computational efficiency of computing-intensivesteps, arrangement of processing units for easy data sharing, andefficient parallel training (e.g., segmenting tensors to evenlydistribute workloads to processors), an architecture that fullyexploits the sparsity of tensors to improve computationefficiency.

The state-of-art AI systems are far from perfection. Robustness,safety, reliability, data privacy, are just some of the mostnoticeable pain points in training and deploying AI systems. Forexample, an AI model trained from a first domain may havenear-perfect accuracy for inferencing in the first domain, butgenerate disastrous inferences when being deployed in a seconddomain, even though the domains share some similarities. Therefore,how to train an AI model efficiently and adaptively so that it isrobust when being deployed in all domains of interest is bothchallenging and intriguing.

As another example, AI systems trained based on training datamay be easily fooled by adversarial attacks. For instance, asecond deep neural network may be designed to compete against thefirst one to identify its weaknesses. The safety and reliability ofsuch AI systems will be critical in the coming years and may beimportant patentable subject matters.

As yet another example, training data in many cases may includesensitive data (e.g., customer data), directly using such trainingdata may result in serious data privacy breaches. This problembecomes more alarming when a plurality of entities collectivelytrain a model using their own training data. Accordingly,researchers and engineers have been exploring differential privacyprotection and federated learning to address these issues.

The content of this article is intended to provide a generalguide to the subject matter. Specialist advice should be soughtabout your specific circumstances.

POPULAR ARTICLES ON: Intellectual Property from United States

Obhan & Associates

Trademarks Comparative Guide for the jurisdiction of India, check out our comparative guides section to compare across multiple countries

Haug Partners

In a June 11, 2021 decision, Yu v. Apple Inc., a Federal Circuit panel issued a precedential decision, with a dissent, upholding the invalidation of patent claims to a digital camera on a motion to dismiss.

Brown Rudnick LLP

There are two dates that must be specified in every trademark/service mark application. They are, the date of first use anywhere and the date of first use in commerce.

Original post:

Patent Protection On AI Inventions - Intellectual Property - United States - Mondaq News Alerts