Home Blog Page 107

Experience Amazing Video Personalizations with Myna by Gan.ai

gan.ai

2022 was the year of generative AI, emerging as one of the most strategic technologies that revolutionized the content creation market. The industry is anticipated to exhibit a CAGR (compound annual growth rate) of over 19.80% in the next five years, as per a report by IMARC Group. Generative AI models utilize advanced machine learning to produce text, images, and videos with simple text inputs. These models substantially impact any organization’s marketing, design, and interpersonal communications. 

Realizing the enormous potential generative AI holds, many startups have tried to ride this wave, but not all have been able to create an impact. Unlike others, Gan.ai was able to harness the power of generative AI and develop something powerful. Gan.ai is a budding Indian company that offers unique video personalization services using Myna, its first commercially available product leveraging generative AI.

To get a deeper understanding of Myna and more insights into Gan.ai’s technology, we interviewed Suvrat Bhooshan, Founder and CEO of Gan.ai. 

The Mind Behind Gan.ai and Myna 

Gan.ai, previously known as Gan Studio, was founded by Suvrat Bhooshan, an artificial intelligence enthusiast.  He is an alumnus of Georgia Institute of Technology, a Bachelor of Science (CS), and Stanford University, Master of Science (CS). Suvrat is a deep learning expert with experience in software development and research engineering at Facebook AI, New York. He specializes in building large transformer models and has been published in several renowned deep-learning conferences like NeuRIPS and ACL. 

Upon returning to India at the end of 2020, Suvrat established Gan Studio in 2021 as an AI-driven personalized messaging service provider. The service created numerous personalized variations of messages using single-input videos. Over the year, Gan.ai advanced to offer more complex video personalizations, like the Zomato ad starring Hrithik Roshan.

The Pretense 

Video communication is becoming more and more popular when it comes to content development. But while addressing a broad audience, generating videos is a cumbersome process that limits customization. Gan.ai was founded with the goal of helping organizations optimize and enhance their video creation process and engage with the audience more effectively with personalized landing pages.

While it may seem to work like the standard deepfake technology, Gan.ai’s fundamental difference lies in personalization and not curation. While deepfake is used to replace the person or likeness with another person, personalization only alters certain aspects of the video. For instance, a food delivery agent (like Zomato/Swiggy) sends a personalized video starring an ambassador who surprisingly addresses you by name and mentions your favorites (based on your recent orders).

Build Dynamic Video Experiences with Myna

Myna is the first brainchild of Gan.ai and is a generative AI tool that helps you build dynamic one-on-one video experiences. Its AI engine enables you to curate breathtaking video campaigns that can be tweaked according to user preferences. It takes three simple steps and only a few minutes to produce hyper-personalized videos with choice parameters. All you have to do is shoot and upload your campaign on Myna, select the words that need to be personalized, and just click on generate! The tool has already captured next-level engagement for brands like Zomato, Swiggy, Bajaj, and Aam Aadmi Party (AAP). Myna has also powered Vivo’s most recent campaign, #vivoVBADay2022

Suvrat believes that Gan.ai has the potential to be the face of generative AI and can advance existing text-to-video generative models. He said, “Gan.ai is at the forefront of generative AI, working on state-of-the-art text-to-video generative models. Today, we have developed the best-in-class LipSync and AudioSync models all in-house and are working towards general domain agnostic video generation simply by describing what you want to showcase.”


You can also see Myna in action on their website.

Advertisement

NVIDIA unveils an early access program for Omniverse Avatar Cloud Engine

NVIDIA unveils an early access program for NVIDIA Omniverse Avatar Cloud Engine (ACE), where developers and teams can build virtual assistants and avatars to enhance existing workflows in business applications.

Omniverse ACE is a suite of cloud-native AI microservice that makes it easier to build and deploy intelligent virtual assistants and digitals humans at scale. It improvises the avatar development process and delivers the AI building blocks necessary to add intelligence and animations to avatars, which are built on any engine virtually and deployed on any cloud. The AI assistants and avatars are designed for organizations across industries to enhance existing workflows and unlock new business opportunities.

Read more: TextBox 2.0, a Python library for applying pre-trained language models to text generation

NVIDIA’s early access program provides access to the pre-release versions of ACE animation AI and conversational AI microservices along with:  

  • 3D animation AI microservice for third-party avatars that uses Omniverse Audio2Face generative AI to make realistic facial animations from audio files. 
  • 2D animation AI microservice, Portrait enables easy animations of 2D portraits or stylized human faces using live video feeds.
  • Text-to-speech micro service leverages NVIDIA Riva TTS to synthesize natural-sounding speech from raw transcripts without any additional information like patterns or rhythms of speech.

With NVIDIA’s early access program, teams and developers can also access tooling, sample reference applications, and supporting resources, which can help them get started.

Advertisement

Conversational AI and metaverse company MeetKai unveils new offerings at CES 2023

MeetKai unveils new offerings CES 2023

The Conversational AI and metaverse company, MeetKai, has unveiled new offerings today at CES 2023, which include MeetKai Reality, MeetKai Metaverse Editor, and MeetKai Cloud AI. 

These offerings are the latest additions to its expanding portfolio of conversational AI technologies, enabling more seamless metaverse experiences available from all web-browser compatible devices without any need for specialized hardware.

MeetKai Reality is a software platform that provides support to users in digitizing physical spaces for creating instant 3-D replicas using merely a phone. The tech can bring spaces and objects from the real world into the metaverse without requiring countless hours of custom 3-D modeling or complicated scanning equipment. 

Read More: Ukraine Pharmacies Accepting Cryptocurrency Payments Via Binance Pay 

MeetKai Metaverse Editor will usher in new ways of metaverse content creation. Unlike other design platforms that struggle with user-friendliness, object physics, or clipping, MeetKai’s new software allows easy virtual space building and construction without complicated tooling. 

MeetKai’s Cloud AI will facilitate another level of immersion in virtual spaces by enabling AI-human interactions with the first cloud offering, which is domain-specific, knowledge-driven Virtual Humans. 

Advertisement

Amazon Confirms to Lay Off 18,000 Employees Under an Uncertain Economy

amazon lay off 18000

CEO Andy Jassy stated that Amazon plans to lay off more than 18,000 employees to reduce costs in an uncertain economy. He added that the company hired rapidly during the pandemic, and such a large workforce is not necessarily beneficial at uncertain times. Amazon doubled its global workforce between 2020 and 2022 by hiring to meet an explosion in demand for deliveries.

The sudden announcement resulted from an employee leaking the information externally. It would not have been revealed until as late as January 18, when the employees to be laid off would have been informed. 

The company’s management, according to Jassey, is ” deeply aware that these role eliminations are difficult for people, and we don’t take these decisions lightly.” Jassy also briefed that Amazon will support those who are adversely affected by extending severance packages, including transitional health benefits and succeeding job placement aid.

Read More: The Universal AI University: India’s 1st AI University

As per Amazon, the layoff is a call for intervention because the company incurred a lot of expenditure across various verticals. The same company has also reached out for a US$8b term loan from certain vendors (undisclosed) to support corporate purposes.


Moreover, the online retailer is anticipating much slower growth as inflation is soaring. Many other companies, like Alphabet, Kraken, Salesforce, Hewlett Packard, and Meta, are also laying off employees to cut costs.

Advertisement

AI to Now Narrate, Apple Reveals AI-Voiced Audiobooks

apple ai audiobooks

In a move that could spell the end for human narrators, Apple has revealed a selection of books using artificial intelligence-based narration. The move is an effort to disrupt the rapidly expanding audiobook business while deepening scrutiny of claims that Apple engages in anti-competitive behavior. 

Since audiobooks have become so popular, technological companies are exploring the domain to have windfalls, especially when the audiobook market is likely to surpass US$35b by 2030. 

Recently, Apple reached out to independent publishers, including those in the Canadian market, as potential partners to develop AI-voiced audiobooks. Authors were to sign a non-disclosure agreement to ensure the secrecy of the project at the time.

Read More: TextBox 2.0, a Python library for applying pre-trained language models to text. generation

Many publishers and authors expressed that the project, if successful, could have a significant effect on the market, while many others were skeptical. David Caron, a renowned co-producer at Canada’s largest audiobook publisher, said, “They’re creating something that is different from the print book, but that adds value as an art form.”

Professional voice actors, however, have not taken Apple AI audiobooks positively as they see it hampers their financial incentive, both as writers and voice agents. On the other hand, if analyzed from a publisher’s point of view, a human-voice audiobook takes weeks and costs over a thousand dollars. AI has the potential to cut both time and costs significantly.

Advertisement

Nuralogix unveils AI app at CES 2023 that uses selfie to determine health-related measurements 

Nuralogix AI app selfie health-related measurements

A startup called Nuralogix unveiled its flagship health and wellness app, Anura, at CES 2023. It takes a 30-second selfie and uses that data to create a catalog of health-related measurements about you. 

The measurements include essential stats such as heart rate, blood pressure, and mental health-related diagnostics such as stress and depression levels. It also includes details about your physical condition, such as body mass index, skin age, and level of risk for things like high blood pressure, stroke, and heart disease, along with biomarkers such as blood sugar levels.

Nuralogix built Anura on an AI trained on data from approximately 35,000 different users. A typical 30-second video image of the user’s face is analyzed to see blood movement around it.  

Read More: Ukraine Pharmacies Accepting Cryptocurrency Payments Via Binance Pay 

According to the company, Human skin is translucent. The light and its respective wavelengths are reflected off different layers under the skin and can be used to reveal blood flow information in the human face.

This is matched against various diagnoses of people using traditional measurement tools and uploaded to the company’s “DeepAffex” emotion AI engine. Anura app users are then essentially “read” based on what the AI ​​is trained to see. The blood flowing in one direction or another, or the color of a person’s skin, can say a lot about that person’s physical and mental state.

Advertisement

Ukraine pharmacies accepting cryptocurrency payments via Binance Pay 

Ukraine pharmacies accept cryptocurrency payments

Pharmacies in Ukraine are accepting digital payments in lieu of the ongoing war, with a significant local pharmacy chain allowing payments in cryptocurrencies such as Bitcoin (BTC). 

ANC Pharmacy, which is one of the biggest pharmacy chains in Ukraine, has collaborated with Binance Ukraine to introduce cryptocurrency payments through the contactless crypto payment service Binance Pay

Starting from January 3, ANC Pharmacy is allowing its customers to go ahead with instant payments while purchasing pharmacy products online, the pharmacy chain announced.

Read More: How SigiloTech Helps SME Clients Scale Operations Using Automation First Approach

ANC Pharmacy has over 1,000 pharmacies across Ukraine and runs its own online pharmacy service. Binance Pay payments will be rolled out in pharmacies in Kyiv initially, according to Binance.

The new payment features will be available at ANC pharmacies, along with ANC Pharmacy-operated pharmacy stores like Kopiyka and Shara. Users must download the Binance application and then go to the ANC.ua website in order to proceed with a crypto payment at ANC pharmacies. 

After selecting a product to order online, users will have the option to pay using the Binance Pay service and pick up the order at a preferred pharmacy.

Advertisement

TextBox 2.0, a Python library for applying pre-trained language models to text generation

A group of researchers from the University of Montreal, Renmin University of China, and Xidian University developed an extended version of the existing text generation package (TextBox 1.0), TextBox 2.0. TextBox 2.0 significantly improves pre-trained text generation models compared to its previous version. It is an up-to-date Python library based on PyTorch that builds a unified and standardized pipeline for applying pre-trained language models to text generation. 

The TextBox 2.0 Python library uses over 45 pre-trained language models that cover 13 tasks and 83 datasets to implement a unified framework for researching text generation. Existing Python libraries fail to support the development of language models in a unified manner as they do not maintain a comprehensive evaluation pipeline for text generation enclosing data loading, training, and evaluation. This is because existing Python libraries are intended to handle a few generation tasks only.

Read more: IIT Patna launches a new BTech course in Artificial Intelligence and Data Science

The TextBox 2.0 library offers a standard method to compare various models and evaluate the generated text. It also provides four rich and influential training methodologies and four pretraining objectives to help users optimize pre-trained language models for text generation. For research purposes, users can either pre-train a brand-new model from scratch or improvise it. This will help them increase the efficiency and dependability of text generation model optimization.

Researchers have released the TextBox 2.0 library on GitHub, where users can quickly learn to install and use it.

Advertisement

SpaceX launched second ‘cryptographically-equipped’ satellite from Cryptosat

cryptosat second satellite crypto2 launch SpaceX
The first launch of 2023 for a SpaceX Falcon 9

Last year in May, Cryptosat — a startup creating satellites that beam cryptographic building blocks down to Earth — had successfully put its first satellite, Crypto1, into orbit with help from SpaceX. Now, Cryptosat’s crypto-satellite network has grown with the launch of its second orbiting satellite, Crypto2. On January 3, the cryptographically equipped device was launched into space onboard SpaceX Falcon 9. 

Cryptosat software was part of the consignment on SpaceX’s maiden voyage in 2023, supporting Cryptosat’s goal of establishing a Trusted Execution Environment in orbit. In addition to Cryprosat, SpaceX launched 114 other tiny satellites for 23 operators into polar orbit.

Comparatively, Crypto2 boasts 30 times more computing power than Crypto1. With the addition of new satellites to the constellation, Cryptosat is trying to turn space into a new “battleground” in the quest for bulletproof cryptography by increasing computer power. 

The launch of Crypto2 is a significant step toward infrastructure expansion from a single satellite in 2022, according to Yonatan Winetraub, co-founder of Cryptosat. “This launch gives us more availability and more powerful specs to support the growing portfolio of use cases in our development pipeline,” says Yonatan.

Read Also: Turkey Announces Blockchain-Based Digital Identity To Login Online Public Services

By isolating computer environments in orbit, where they cannot be as easily targeted, Cryptosat attempts to enhance cryptographic security for sensitive operations. The startup ensures that there has been no outside interference with the computing environment before it is deployed into space by using its own satellites that were designed from the ground up.

With the launch of Crypto1, Cryptosat sought to make blockchain applications more accessible by offering a physically inaccessible, tamper-proof platform. The International Space Station, or ISS, had previously served as a testing ground for the satellite’s technologies to investigate data security applications in space.

Last month, Cryptosat published an API tutorial with a simulated satellite trajectory. This was crucial because now developers can acquire an idea of how to communicate with Cryptosat’s low earth orbit (LEO) satellites in space using the Cryptosim software.

Cryptosat is not the only company that has experimented with blockchain technology’s applications in outer space. As part of SpaceX’s CRS-19 commercial resupply service mission, SpaceChain, a space blockchain company with its headquarters in Singapore, sent its blockchain hardware wallet technology to the International Space Station (ISS) in 2019. Earlier that year, as part of its Kick-start Activity program, the European Space Agency (ESA) gave SpaceChain funding to explore and uncover commercial use cases for its blockchain-based satellite technology. SpaceChain had also collaborated with PricewaterhouseCoopers in France to research its space applications.

Advertisement

Samsung unveils its Bespoke AI Wall Oven with built-in camera 

Samsung Bespoke AI Wall Oven

Samsung has unveiled one of the new smart kitchen appliances, its Bespoke AI Wall Oven, ahead of the CES 2023, featuring a built-in camera, burn detection AI technology, and even allowing livestream view on social media.

The AI Pro Cooking algorithm in the appliance can recognize up to 80 dishes (up to 106 in Europe). It can recommend the appropriate time, temperature, and mode for cooking. It can also send notifications to help prevent the food from burning.

The oven also has a push-to-open door for a sleeker appearance, which also enables the convenience of nudging the door open with one’s shoulder when hands are full. The Bespoke AI Oven is currently available in Europe and is expected to launch in North America in the third quarter. 

Read More: How SigiloTech Helps SME Clients Scale Operations Using Automation First Approach

Samsung also launched three new Bespoke refrigerator models, viz., the Top Mounted Freezer Refrigerator, the 4 Door Flex Refrigerator with Family Hub, and the Side-by-Side Refrigerator.

Inside the refrigerators, the AI-powered ViewInside cameras help to keep track of inventory and can also analyze food images and labels.

Advertisement