Eros Investments and Wipro to evolve AI & ML based content localization solution

By

-

June 20, 2022

Eros Investments and Wipro content localization solution

Eros Investments has announced a signed alliance agreement with Wipro to scale and evolve the Machine Learning (ML) and Artificial Intelligence (AI) based content localization solution.

Eros Investments is the entertainment, media, and technology portfolio of ventures such as Eros Now, Eros Media World, Xfinite’s Mzaalo, and others. At the same time, Wipro is a technology service and consulting company.

The AI and ML-based solution will automate the time-consuming manual content localization process of dubbing and subtitling. It is expected to be done with near human-level accuracy, driving significant time and cost savings for global media organizations, direct-to-consumer over-the-top (OTT) streaming platforms, and post-production.

Wipro’s Vantage solution is an ML/AI-powered content intelligence platform that uses Google Cloud’s Translation AI suite of services to generate computer-generated voices from Text to Speech and Speech to Text models. The platform also includes emotion tagging, voice cloning, and speed syncing in various languages. Wipro’s Vantage also assists in extracting intelligence/metadata from multiple forms of content, including audio, video, images, printed text, and more.

In collaboration with Wipro’s technology team, Eros Investments’ data science experts will leverage the platform.

The first phase of subtitling or automated translation will be available in numerous languages, including French, Spanish, English, Arabic, Malay, Mandarin, Bahasa, Tamil, Hindi, Telugu, and Bengali. The ‘use cases’ from these languages will later be used to train models to develop the solution in other languages.

According to Victor Morales, Managing Director, Global Systems Integrator Partnerships, Google Cloud, Eros Investments and Wipro’s content localization service, when collaborating with Google Cloud’s machine learning capabilities, will avail the customers across the media sector the functionality they need to deliver exceptional viewing experiences to their audiences everywhere.

Fujitsu and Intel display Real-Life AI- Solutions at AI Summit 2022

By

Disha Chopra

-

June 20, 2022

AI Summit London: The info-tech company Fujitsu and processor chips manufacturer Intel showcased their AI solutions on display at the AI Summit in London. The demonstration showed how convenient integrating AI technologies into daily life is.

Intel displayed its Voice Creation Software, 3D athlete Tracking Wearable, and Food Redistribution Network to feed those in need. The three technologies demonstrate the best tech that genuinely impacts people’s lives, with the athlete tracking model and food redistribution being the more practical AI benefits.

“There’s a lot of talk in the industry about AI, and we wanted to highlight where it’s making a real difference. They’re powerful stories with relatable AI, which we thought people would get immediately,” said Chris Feltham, a tech specialist at Intel.

He also said, “The best AI is the one that you don’t even realize is there. Our goal is to be as accessible as possible and overcome any skills gaps existing in the AI and make it as easy as possible.”

Fujitsu displayed its Quantum Supercomputer and Digital twin technologies. The company co-developed the supercomputer Fugaku with its optimization system Digital Annealer.

Fujitsu’s CTO Matthew Chase, said, “The supercomputer can determine the potential outcome of something depending on the pattern it recognizes. With something like traffic flows, we can use AI to spot and monitor human behavior to identify people in trouble and preemptively prevent any problems.”

Fujitsu is also working on digital twin technologies to provide a “quantum harness” while focusing on interoperability and synergy. The idea is to provide a platform for quantum scientists to practice on.

Google AI Launches LIMoE, a Large-Scale Architecture for Pathways

By

Disha Chopra

-

June 20, 2022

Google AI launched a breakthrough new technology called LIMoE as a step towards its far-reaching goal of an AI architecture known as Pathways. Pathways is a single model AI architecture that can accomplish multiple tasks. It encapsulates Google’s research goal to use sparse models that handle text and images simultaneously.

Read all about LIMoE (the Language Image Mixture of Experts), the first large-scale architecture that processes both images and text using a sparse mixture of experts, which achieves high performance with much less compute than other top methods. → https://t.co/eKOeT3UaKn pic.twitter.com/7mxUscv1Js
— Google AI (@GoogleAI) June 9, 2022

LIMoE stands for Learning Multiple Modalities with One Sparse Mixture-of-Experts Model. It is not the only architecture that can multi-task. Nevertheless, what sets it apart from others is the use of sparse models. Sparse models distinguish as one of the most promising future approaches to deep learning. They are different from the ‘dense’ models in that the sparse-model routes specific tasks to other “experts” in the network instead of using dependent computation instead of conditional computation.

Google AI team has presented a sparse mixture of experts in “Multimodal Contrastive Learning with LIMoE: the Language Image Mixture of Experts.” the technology analyzes words and images simultaneously with sparsely activated experts.

It outperforms other dense multimodal models and techniques in zero-shot image categorization. LIMoE can learn to handle a range of inputs and scale them up because of its sparsity.

In the announcement, Google experts explained how LIMoE works. “There are also some clear qualitative patterns among the image experts — e.g., in most LIMoE models, there is an expert that processes all image patches that contain text. …one expert processes fauna and greenery, and another processes human hands.”

Maruti Suzuki invests ₹2 Cr in AI startup Sociograph Solutions

By

Dipayan Mitra

-

June 20, 2022

India’s largest automobile manufacturing company Maruti Suzuki India said on Friday that it had invested Rs 2 crore in Sociograph Solutions, an artificial intelligence (AI) firm.

This new investment is part of the company’s MAIL program, which aims to encourage emerging mobility entrepreneurs. Maruti acquired a 12.09 percent share in the firm, claiming it is the first investment made under the Maruti Suzuki Innovation Fund.

Maruti intends to leverage SSPL’s Dave visual artificial intelligence (AI) platform. AI, to improve its clients’ digital sales experience.

According to the company, to improve the consumer experience, the Dave.AI technology will provide 3D visualization. Since 2019, Maruti has been implementing the ambitious Mobility & Automobile Innovation Lab (MAIL) initiative to strengthen the country’s mobility startup ecosystem.

MSIL Managing Director & CEO Hisashi Takeuchi said, “Our investment in SSPL demonstrates our resolve towards improving business metrics using contemporary technology.”

He further added that the Maruti Suzuki Innovation Fund was established with the goal of investing in early-stage firms that are part of Maruti Suzuki programs.

According to Dave.AI’s CEO, Sriram PH CTO Ananth, the partnership with Maruti Suzuki greatly benefits the firm by not only proving its concepts but also learning and assimilating the skills necessary to scale up operations in a sustainable manner.

They said, “Post our collaboration with Maruti Suzuki under the MAIL program, we registered 300 percent growth in revenues and are on track to achieve USD 1 million annual revenue milestone this financial year.”

Snowflake invests in Domino Data Lab to provide deeper integrations

By

Sahil Pawar

-

June 20, 2022

The venture capital arm of the data cloud company Snowflake has announced that it has invested in the Series F of Enterprise MLOps platform Domino Data Lab. The latter also revealed the presence of Snowflake-powered capabilities in Domino 5.2.

The investment on behalf of Snowflake Ventures will allow both the companies to co-develop deeper Domino-Snowflake product integrations. Through the power of Domino’s platform and Snowpark, the developer framework for Snowflake, the companies will be able to deliver an end-to-end enterprise data-science lifecycle solution on a single shared data and deployment platform.

With Domino, data science teams can easily embed model-driven applications throughout the enterprise, thus leveraging Snowflake’s Data Cloud. The partnership between Domino and Snowflake boosts the MLOps lifecycle by eliminating workarounds caused by inflexible tools and complex processes, automating non-data science tasks, and avoiding DevOps tensions between IT and data scientists.

Luca Foschini, Chief Data Scientist at Evidation, said that Domino and Snowflake enable their data science team to connect to person-generated health data (PGHD) seamlessly and prototype rapidly within a secure environment.

The partnership has strengthened Domino’s position as the industry’s most open and flexible Enterprise MLOps platform. This is due to the exceptional capabilities across each stage of the MLOps life cycle.

Integrating Snowpark will bridge the gap between IT, data scientists, and the business by allowing direct execution on Snowflake as the common platform. The new features will be available this month.

Christian Kleinerman, Snowflake SVP of Product, said that the investment with Domino solidifies their partnership and joint commitment to help the enterprise customers leverage the power of the Data Cloud through the adoption of machine learning models.

Canada announces new Laws for better Privacy Protection

By

Dipayan Mitra

-

June 19, 2022

The government of Canada announces new laws as a step to strengthen the country’s privacy protection policies and build trust among citizens regarding the digital economy.

The Government of Canada is dedicated to ensuring Canadians can benefit from cutting-edge technology while knowing that their personal information is safe and secure and that companies are operating responsibly.

The new bill named C-27 was introduced by the country’s Minister of Innovation, Science and Industry, François-Philippe Champagne, and Minister of Justice and Attorney General of Canada David Lametti. .

According to the government, the Digital Charter Implementation Act 2022 would considerably strengthen Canada’s private sector privacy law, set new standards for the responsible development and use of artificial intelligence (AI), and continue to advance Canada’s Digital Charter implementation.

David Lametti said, “We are committed to protecting the personal information of Canadians while encouraging responsible innovation for the public good. The Digital Charter seeks to bring essential online protections for individuals into the 21st century.”

He further added that their government is also updating the Privacy Act, Canada’s federal public sector personal information protection act, to ensure that the country’s privacy regulations keep up with technological development and reflect developing Canadian values.

The Consumer Privacy Protection Act, the Personal Information and Data Protection Tribunal Act, and the Artificial Intelligence and Data Act are all proposed legislation within the Digital Charter Implementation Act, 2022.

The government mentioned that the proposed Consumer Privacy Protection Act would address the demands of Canadians who rely on digital technology while also responding to criticism of prior proposed laws.

“In today’s economy, Canada’s competitiveness depends on our ability to use digital innovation to harness the power of data. Safety and trust must be the foundation of this new digital economy,” said François-Philippe Champagne, Minister of Innovation, Science, and Industry. Champagne also mentioned that they guarantee Canadians can trust when and how their information is used by adopting the Digital Charter Implementation Act, 2022.

NVIDIA DRIVE Orin Powers Intelligent Robo-01 Concept Vehicle by JIDU

By

Disha Chopra

-

June 19, 2022

The EV startup JIDU announced NVIDIA DRIVE Orin for its intelligent Robo-01 concept vehicle during the ROBODAY event. The EV company is backed by Baidu, a tech giant, whose metaverse platform showcased the ROBODAY presentation for Robo-01.

🚘JIDU unveiled its first concept production car #ROBO-01 today 😎 Integrating the leading intelligent driving capability powered by Baidu and a cutting-edge futuristic design, ROBO-01 is set to lead the way to an intelligent car era driven by AI. More at https://t.co/AYnUFZY0eh
— Baidu Inc. (@Baidu_Inc) June 8, 2022

The robocar would feature cutting-edge AI capabilities inspired by robots and powered by the high-performance NVIDIA platform. JIDU provided a glimpse of the sleep SUV and the upcoming lineup.

The cars would be capable of level 4 autonomous driving, driverless parking, and safe & sound operation at highway speeds. Robo-01 also features lidar sensors that can retract and extend from the hood. Besides, it will come with a retractable steering wheel that retracts under the dashboard when you switch to autonomous driving.

The concept vehicle would have the ability to enhance by appending new intelligent features throughout its life. NVIDIA DRIVE Orin will come into the picture when these updates are delivered over the air. This is why JIDU has debuted with a software-defined vehicle architecture built on the AI compute. Robo-01 comes with two NVIDIA DRIVE Orin systems-on-a-chip (SoC). These chips will provide ample computing for autonomous driving.

The two DRIVE Orin SoCs will aid in achieving redundancy and variation necessary for in-cabin AI technology. The lineup will be jam-packed with features and the scope to add even more. JIDU has already planned to launch a limited-production version of the concept later in the year.

The automaker plans to unveil the design for the same at Guangzhou Auto Show in November 2022.

It plans to start taking orders for Robo-01 in 2023 while the deliveries are scheduled for 2024.

Miso Robotics partners with Americana Restaurants at Wimpy in Dubai Mall

By

Dipayan Mitra

-

June 19, 2022

Miso Robotics Americana Restaurants Dubai

Artificial intelligence-powered robots developing company for the restaurant industry Miso Robotics partners with Americana Restaurants at Wimpy in Dubai Mall to deploy its Flippy 2 frying robots.

Americana Restaurants is a renowned F&B operator and master franchisee known for delivering famous worldwide brands to the Middle East and North Africa, such as KFC, Pizza Hut, Hardee’s, Krispy Kreme, and TGI Friday’s.

Miso Robotics claims that its Flippy 2 can accomplish the duty of a complete fry station on its own, streamlining operations in quick-service restaurants (QSRs). The robot has demonstrated its ability to optimize operations at QSRs and will be integrated and expanded across more Americana Restaurants locations in the coming months.

Melvin Michael, Brand Director of Wimpy, said, “Americana Restaurants’ partnership with Miso Robotics marks a significant milestone for the company as we continue to strive to address shifting customer behaviors and preferences through state-of-the-art, innovative solutions.”

He further added that as the first QSR operator in the MENA region to integrate robots in its kitchens, Americana Restaurants displays its dedication to providing consumers with new eating experiences at their most iconic brands. According to the company, Flippy 2 will provide a customizable robotic fry station to Wimpy’s kitchen, resulting in a reduction in order-to-delivery time, an increase in food uniformity, and a better working environment for human team members.

Mike Bell, CEO of Miso Robotics, said, “In the past year, we’ve announced partnerships with some of the largest brands in the U.S., and that momentum has led to us being pulled internationally to introduce our technology on a global scale.”

Recently, Bakery chain Panera Bread also started to deploy Miso Robotics’ new automated coffee brewing system, CookRight Coffee, at its outlets. The one-of-a-kind robot ensures a quality cup of coffee and a more efficient team member experience by monitoring indicators such as volume, temperature, and time data and combining it with predictive analytics.

TigerGraph launches new machine learning modeling tool

By

Sahil Pawar

-

June 19, 2022

TigerGraph launches machine learning modeling tool

TigerGraph, a California-based company, has launched a new machine learning capability explicitly designed to boost the development and improve the accuracy of data science models.

The graph database vendor, founded in 2012, announced the TigerGraph ML Workbench in a preview during the Graph + AI Summit, the spring edition of the biannual open conference hosted by TigerGraph, virtually.

The new capability will act as a platform for developing and deploying machine learning models with the latest graph technology. It works with leading third-party machine learning platforms.

The platform by TigerGraph uses graph technology that enables data points within databases to simultaneously connect to multiple other data points, instead of just one data point at a time as in the case of a traditional relational database. Users can more easily discover relationships between data by connecting to multiple data points simultaneously. This results in speed and accuracy, days the vendor.

The tool is compatible with TigerGraph 3.2 and versions succeeding that. It can be deployed as either a fully managed cloud service or on-premises and will be available in June 2022.

More typical use cases for graph technology include the development of social media platforms and fraud detection, which rely heavily on discovering relationships.

TigerGraph ML Workbench aims to make data science, explicitly developing machine learning models, faster, easier, and more accurate. According to Victor Lee, vice-president of machine learning and AI at TigerGraph, Graph technology is boosting the analytics process and enabling users to reach insight and action more accurately than they could with relational databases.

Veritone acquires AI-powered Voice Creator VocaliD

By

Dipayan Mitra

-

June 19, 2022

Machine learning and artificial intelligence company Veritone acquires AI-powered voice creator firm VocaliD. Neither company provided any information regarding the valuation of this acquisition deal.

Veritone Voice can now offer the capacity to oversee and manage the entire voice generation lifecycle for efficiency and scale, using the expertise of VocaliD.

Integrating VocaliD’s powerful voice models, patented vocal correction, and audio mixing features with Veritone Voice expands the enterprise-grade solution while lowering complexity, cost, and administrative time.

Veritone Voice, based on Veritone’s aiWARE, comes with the unique ability to effortlessly integrate third-party AI models, allowing the VocaliD model to be used immediately.

Ryan Steelberg, Veritone President and Co-founder, said, “Veritone Voice brings enormous value to our enterprise clients and IP owners of all sizes. With such a large addressable market, Veritone is able to meet consumer demands for content, at scale, with localization, and without sacrificing quality.”

He further added that he is delighted to bring in VocaliD to enhance their Voice service and deliver even more capabilities to current and potential clients from both a people and a technology standpoint.

United States-based voice artificial intelligence company VoacliD was founded by Rupal Patel in 2014. Using machine learning techniques, proprietary voice mixing algorithms, and a crowdsourced Voicebank, VocaliD’s award-winning technology delivers high-quality, natural-sounding voices in hours. To date, the company has raised more than $3 million from multiple investors over several years.

Rupal Patel said, “We are thrilled to join forces with Veritone, to bring our vision of leveraging AI voice to create more equitable, just, diverse, inclusive, expressive, and engaging user experiences to life.”

Patel further added that they are well-positioned to set industry norms and establish new monetization models that benefit all stakeholders, given their common commitment to responsible and ethical AI.