Researchers from Disney have announced a new aging/de-aging tool, Face Re-aging Network (FRAN), a neural network that can make actors look older or younger without needing weeks of complex visual effects work.
The technology will enable TV or film producers to re-age actors using an automated, cost-effective, and less time-consuming process than previous methods.
Disney presented the research in a paper called “Production-Ready Face Re-Aging for Visual Effects” and submitted it to be included at the 2022 ACM SIGGRAPH Asia conference to be held in December.
Traditionally, when special effects staff on a film or video production need to re-age the actor, they usually either use a 3D modeling and 3D scanning process or a 2D frame-by-frame digital retouching of the face using tools like Photoshop. This process can take weeks, depending on the length of the work.
In contrast, Disney’s new AI technique automates the process. Disney calls it the first production-ready, practical, and fully automatic method for re-aging faces in video images.
To build FRAN, Disney researchers randomly created thousands of samples of synthetically aged faces between ages 18-85 using StyleGAN2. With that training data in hand, FRAN learned general principles about how a person’s appearance changes with age. Now that training is complete, it can apply those aging principles to a real actor in motion, frame by frame.
AWS re:Invent is the biggest and the most comprehensive cloud computing event hosted by Amazon Web Services, a leading cloud service provider. The annual event features several announcements, innovations, and training opportunities for the cloud computing community and has been one of the most transformative conferences in the last 10 years. AWS re:Invent 2022 commenced on November 28 in Las Vegas. Adam Selipsky, CEO of AWS, initiated the event with a water sustainability pledge to return more water to communities than its usage in cloud operations by 2030. Others, including Swami Sivasubramanian (VP of AWS Data & ML), Dr. Werner Vogels (VP and CTO at AWS), and other executives, are also speaking on new product and innovation launches.
This article highlights some AWS re:Invent 2022 announcements made during the global conference.
Top AWS re:Invent 2022 Announcements
Have a look at some of the major announcements made at AWS re:Invent 2022.
1. Announcements on Compute and Cloud Management
AWS introduced large-scale spatial simulations with SimSpace Weaver at the AWS re:Invent 2022. SimSpace Weaver is a novel computing service running real-time spatial simulations simultaneously across several Amazon EC2 instances. Using this service, developers can overcome computing and memory constraints while running jobs, as quoted by Selipsky.
AWS has also introduced ENA (Elastic Network Adapters) Express for improved network latency and performance on Amazon EC2 instances. ENAs are built on top of AWS’ Scalable Reliable Datagram (SRD) Protocol and reduce latency by up to 50% while increasing network bandwidth.
The event also witnessed Wipro’s launch of Skill Guild in association with AWS. Skills Guild is a comprehensive cloud-based program that will help enterprises to build cloud skills with AWS learning opportunities.
Informatica, a California-based cloud data management company, also introduced a few new offerings at AWS re:Invent 2022 event. The company introduced Informatica Data Loader (for Amazon Redshift console), Informatica Data Marketplace (for AWS data exchange), and Informatica INFACore (to support AWS SageMaker Studio). These services will help departmental users, developers, and data scientists across all skill levels.
AWS also introduced the preview of Amazon CloudWatch Internet Monitor, a new monitoring service for customers to diagnose internet performance issues and generate performance metrics.
2. Partnerships and Collaborations at AWS re:Invent 2022
AWS and TCS (Tata Consultancy Services) have collaborated and launched the TCS Quantum Computing Lab on AWS to enable faster exploration, development, and testing of business solutions in the adoption of quantum computing.
Compass UOL, a technology service provider, has partnered with AWS to build an AWS-owned and operated business unit to accelerate IT modernization in both private and public sectors.
AWS is also working with Siemens to help provide its industrial software into the AWS cloud to help startups in scalability and performance.
3. Announcements catering to the Healthcare Industry
Several AWS re:Invent 2022 announcements show AWS’ work in the healthcare industry to provide specialized tools for better patient care and treatment.
AWS is also exploring scalable data integration from multiple environments. It has announced AWS Clean Rooms for enhanced data collaboration and analysis, enabling companies to work together on massive datasets.
To help mitigate the risks associated with supply chain management, AWS has introduced a preview of the AWS Supply Chain, a cloud-based application that reduces cost and increases resilience.
The AWS re:Invent 2022 event introduced Amazon Omics, a purpose-built service that stores, queries, and analyzes genomic and biological data. The new service will help bioinformaticians, scientists, and researchers work on medical data and advance scientific discoveries.
4. Some other general announcements
AWS is a founding member of the newly-launched Open Cybersecurity Schema Framework and released a preview of the Amazon Security Lake service to enhance the operationality of the cybersecurity framework.
The AWS re:Invent 2022 event also witnessed the reveal of EKS Protection for Amazon GuardDuty, a service that enables users to detect and respond to security threats online at scale.
In addition, the company announced a preview of the serverless Amazon OpenSearch Service. Giving the existing service an option to go serverless simplifies the process of running jobs at a large scale and minimizes the need to configure, manage, or scale OpenSearch clusters.
AWS also announced Amazon DataZone at AWS re:Invent 2022, a new data management system that will bring employees closer to organizational data while maintaining confidentiality. The system is highly integrable with standard AWS tools like Redshift. Selipsky highlighted that finding the right balance between access and control of enterprise data is crucial but different for each enterprise. Hence, DataZone will be accessible, guided, and customizable.
5. A redesigned user interface for Amazon SageMaker Studio
Amazon’s integrated development environment (IDE) SageMaker Studio received a new, user-friendly interface. The company announced the redesign at AWS re:Invent 2022 to introduce a new navigation menu highlighting clear entry points into SageMaker’s capabilities. The navigation is a collapsible bar backed by an architecture that reflects standard ML workflows. The redesigned interface is available in all AWS regions where SageMaker is available.
6. Amazon Redshift Multi-AZ (preview) support for RA3 clusters announced at AWS re:Invent 2022
Redshift introduced a preview of Multi-AZ deployments to support running data warehouses in multiple AWS availability zones (AZ) simultaneously. The new setup will help clients with business-critical analytical applications achieve the highest availability and resilience to AZ failures. The preview is currently available in the following AWS regions: US East (N. Virginia), US East (Ohio), US West (Oregon), Asia Pacific (Tokyo), Europe (Ireland), and Europe (Stockholm).
7. New-Trusted Language extensions for PostgreSQL on Amazon RDS and Aurora
PostgreSQL is one of the most reliable open-source databases, and numerous enterprises use it because of its convenience and extensive design. Amazon already offers as many as 85 PostgreSQL extensions in the PostgreSQL-compatible edition of Amazon Aurora and Amazon Relational Database Service. However, at the AWS re:Invent 2022 event, the company announced the general availability of Trusted Language Extensions for PostgreSQL. It is an open-source development toolkit for generating PostgreSQL extensions and running them safely. The toolkit will also give developers control over who can use these extensions via permissions models for running them.
8. AWS Glue Data Quality Preview unveiled at AWS re:Invent 2022
AWS Glue has announced a preview version of AWS Glue Data Quality, an automated capability that measures and monitors data lakes and pipelines. It is serverless and allows scalable data integration for more efficient data flows from multiple sources. Extract, transform, and load (ETL) jobs on data pipelines using AWS Glue, which can also set data quality standards and actions. These rules can stop “bad” data from accumulating in data lakes and warehouses. The preview is available for select AWS regions, including US East (Ohio), US East (N. Virginia), US West (Oregon), Asia Pacific (Tokyo), and Europe (Ireland).
9. AWS AI Service Cards
AWS re:Invent 2022 also witnessed the launch of a new resource that enhances the transparency of artificial intelligence-based AWS services. AWS AI Service Cards are a kind of responsible AI documentation that acts as the point of reference for clients to find information on use cases, limitations, design choices, and deployment methods for AWS’ responsible AI services. The service will include three cards: Amazon Rekognition – Face Matching, Amazon Textract – AnalyzeID, and Amazon Transcribe – Batch.
10. General availability of Amazon DocumentDB Elastic Clusters
AWS announced the general availability of Amazon DocumentDB Elastic Clusters, a new Amazon DocumentDB cluster that enables you to scale document databases elastically. Many of these clusters’ management features, such as Multi-AZ support, AWS CloudWatch integration, automated patching, and snapshot backups, are shared by Amazon DocumentDB instance-based clusters.
11. Preview of AWS Application Composer
The third day of AWS re:Invent 2022 also unveiled a preview of Application Composer, a new service that helps to simplify and accelerate architecting and building serverless applications. It can develop and maintain deployment-ready infrastructure, levering one of the first IaC (infrastructure-as-code) services. The preview is available in the standard select AWS regions.
12. Amazon CodeCatalyst Preview was also announced at AWS re:Invent 2022
AWS introduced a preview of Amazon CodeCatalyst, a unified service streamlining software development and delivery on AWS. The entire project experience offered by CodeCatalyst combines the tools required to plan, code, build, test, and deploy apps. With CodeCatalyst, software development teams can automate many complexities, like setting up tools, CI/CD, integration, etc., and focus on direct applications. The service can also browse a collection of blueprints for standard projects and then choose one to set it up automatically. Users can sign up here and try it on their own.
13. AWS Ambit Scenario Designer
The new AWS Ambit Scenario Designer is a set of tools to accelerate the generation of 3D content for simulations of robots, autonomous vehicles, and other near-real-time 3D applications. Using this new service announced at AWS re:Invent 2022, developers can experiment with simulation scenarios in a fraction of the time taken by traditional methods.
14. Amazon’s Distributed Computing Manifesto
AWS re:Invent 2022 ended with Dr. Werner Vogels talking about quantum computing, the challenges of quantum computing, and how AWS is working on the same. He announced Amazon’s Distributed Computing Manifesto, an age-old canonical document that helped transform Amazon’s e-commerce platform. Dr. Verner published the manifesto in November to re-iterate the challenges that the company is yet to face and where it is headed.
Samsung said it plans to hire about 1,000 engineers from IITs and leading engineering institutes to work on newer technologies at its R&D institutes throughout India at a time when tech giants are laying off thousands of employees across the globe.
The new workforce will work next year at Samsung R&D Institute-Bangalore (SRI-B), Samsung Semiconductor India Research in Bengaluru, Samsung R&D Institute-Noida, and Samsung R&D Institute-Delhi.
The new hires will work on cutting-edge technologies such as the internet of things (IoT), machine learning, image processing, cloud storage solutions, big data, business intelligence, predictive analysis, and system-on-a-chip (SoC), among others.
Samsung will recruit engineers from various streams, such as information technology, electronics, instrumentation, embedded systems, computer science and allied branches, and communication networks.
In addition, the company also plans to hire from fields such as mathematics, computing, and software engineering. This hiring season, Samsung R&D centers will employ around 200 engineers from IITs. They have also offered more than 400 pre-placement offers to students from the IITs and other leading institutions.
OpenAI has introduced a new conversational language model, ChatGPT, for optimizing dialogue. Having trained similarly to InstructGPT, the language model works on the dialogue format to answer questions, challenge premises, and reject derogatory requests.
OpenAI’s GPT (generative pre-trained transformer) language models have been doing well since the GPT-1 and GPT-2 came out. The first two models focused on using accurate datasets and adding parameters to make the model more precise and robust. With GPT-3, the company shook the industry by training it on 570GB of information.
However, with all language models, there is a risk of them generating toxic and inappropriate results.
It has also raised concerns about its potential to spread misinformation and generate harmful language. To mitigate these risks, the ChatGPT Detector can help identify and flag such content, allowing for further review and correction. As OpenAI continues to develop more advanced language models, tools like the ChatGPT Detector will play an increasingly important role in promoting the responsible and ethical use of natural language processing technology.
However, OpenAI is attempting to make language models safer and more ethical. ChatGPT is a fine-tuned version of the GPT 3.5 series and has been trained using Reinforcement Learning from Human Feedback (RLHF). The initial model was trained under supervision, where human AI trainers gave conversational inputs.
OpenAI has provided a free preview for people to try ChatGPT and give feedback. For instance, when given a command to suggest birthday party ideas, ChatGPT outputs:
Try it here after creating an OpenAI account if there is not any existing one.
While ChatGPT is an advancement for conversational AI, it is still bound to have some foundational limitations. The language model sometimes writes “nonsensical answers,” a problem with all language models. Additionally, ChatGPT is sensitive to tweaks to the input and is “excessively verbose.” While OpenAI has undertaken all efforts to prevent model toxicity, it is prone to inappropriate usage as it might respond to harmful instructions too.
OpenAI launched a new model in the GPT-3 family of AI-enabled large language models, called Davinci-003, on Monday. Davinci-003 ensures high quality, more extended output, and better instruction-following.
Soon after the launch of the Davinci-003, people discovered that Davinci-003 could generate rhyming songs, limericks, and poetry, which GPT-3 could not do previously.
I posted about how image-generating AI has gotten exponentially better in the last month.
Well, a new text model was released for GPT-3 today. AI can now write rhyming poems. And acrostics. And limericks. And explain how a candy-powered FTL drive can help me escape from otters. pic.twitter.com/vBAroN2SUv
However, OpenAI has not yet commented anything on the rhyming songs and poetry of Davinci-003.
According to OpenAI, Davinci-003 can perform all the tasks that the other models can and even need fewer instructions. It works well with tasks requiring in-depth knowledge of the subject, like summarizing texts for specific audiences and creative content development. As Davinci-003 produces the best results for tasks that need an in-depth understanding of content, it will require more computing resources. Therefore, Davinci-003 costs more per API call and is also a bit slower than the other models.
Davinci-003 excels at understanding the intent of the text. For example, it is good at providing solutions to various logical problems and explaining the character’s motivation. Davinci-300 has been capable of solving most of the challenging AI problems containing cause and effect.
On Monday, Shira Greenburg, the Chief Economist at Israel’s Ministry of Finance, published a list of recommendations for regulating digital assets.
The list of recommendations consists of new regulatory infrastructure, legislation to authorize and control the issuance of backed digital assets like stablecoins, and facilitating financial services. The recommendation also included the requirement for a law to provide the Bank of Israel the authority to oversee digital assets with significant stability or monetary effect.
The recommendation consists of allowing the payment of taxes on cryptocurrency held outside Israel with the central bank. It will also create an inter-ministerial committee to control the regulation of decentralized autonomous organizations (DAOs).
Shira Greenburg suggested that the supervisor of the Financial Service Providers must be given more authority to look after the licensing rules and build a more comprehensive taxation framework for the sale and purchase of digital assets. She also mentioned that Israel should implement more security standards on trading platforms, crypto issuers, and funds generated from digital assets. Read more: Brazil legalizes crypto transactions
Brazil’s Chamber of Deputies approves a bill to regulate the crypto industry on Tuesday. The Senate approved the bill in April 2022, and it was stucked in the Chamber of Deputies. The bill will now need the approval of the executive branch to become law.
The bill is drafted by deputy Aureo Riberio, establishing a new crime of fraud consisting of virtual assets, with a penalty of 2 to 6 years and a fine. It also specifies the creation of a ‘virtual service provider’ license that is to be requested by crypto firms and other companies.
According to the bill, companies can have 180 days to accommodate the new rules before the law is enforced. The bill also says that the Brazilian Securities and Exchange Commission (CVM) will regulate crypto assets securities. In contrast, other digital assets not falling under the crypto assets will fall under the responsibility of another body, which the executive branch will appoint. As per the report, the central bank is expected to be selected as another body by the executive branch.
However, Brazil is not the first nation to have bills regulating crypto transactions. Besides Brazil, other countries, such as China, Cuba, Bolivia, Iran, India, and more, have crypto transactions and other crypto-related activities that are legal and restricted.
A Japanese organization called ispace was supposed to launch its Hakuto-R lunar lander on 30th November. But according to the recent report of ispace, the launch of the Hakuto-R lunar lander is postponed to 1st December. If this ispace’s mission is successful, the Hakuto-R lunar will be the first private spacecraft funded by a private organization to land on the lunar surface.
According to ispace, the lander would touch down in the Atlas Crater by April 2023 on the visible side of the moon. Hakuto-R was one of the five finalists of the international Google Lunar XPrize competition, which was a challenge to land a rover on the moon before a 2018 deadline.
However, Hakuto-R is not the first spacecraft to enter the lunar surface. In 2019, the Israeli organization, SpaceIL attempted to enter the lunar surface but could not land successfully. SpaceIL experienced a fatal engine flaw and crashed into the lunar surface during the landing attempt.
Takeshi Hakamada, the CEO of ispace mentioned that they would like to position Ispace as an international bridge between the US and other countries. ispace now has contracts with NASA and the European Space Agency for upcoming missions to land on the moon and gather lunar water and dust samples.
ispace’s first mission, M-1, is to launch a SpaceX Falcon 9 rocket from Cape Canaveral, Florida. The lander will be carrying a small rover for the United Arab Emirates’s Mohammad Bin Rashid Space Center, a miniature two-wheeler robot for the Japan Aerospace Exploration Agency (JAXA), a camera, and a flight computer prototype for Canadian companies. If Ispace succeeds, it would be the first craft from Japan and United Arabs to visit the lunar surface.
The central government drafted a bill in 2018 to ensure digital privacy for individuals and businesses relating to their data and cement trust between people and entities processing data. The Data Protection Bill, as it was called, was initially proposed in 2018 and tabled by the Ministry of Electronics and Information Technology. The Bill introduced several guidelines for protecting personal data and proposed the establishment of the Data Protection Authority of India.
The initial draft of the PDP Bill provided a robust set of provisions for both domestic and international organizations involved in data processing and other factors related to anonymized data. However, the drafters still welcomed recommendations to make the bill more practical. After considering all recommendations suggested by the cabinet, the revised Personal Data Protection Bill came out in 2019.
Some key provisions that the revised Bill aimed to provide:
Specification of the flow and usage of personal data.
Ensuring fundamental rights of those whose personal data are processed.
Creation of a framework for organizational measures in the processing of data.
Laying down norms for social media intermediaries.
Remedies for unauthorized and harmful processing.
The Bill faced criticism from Justice B. N. Srikrishna, who drafted the 2018 Bill. He said the revisions could turn India into an “Orwellian State.” The term describes a societal condition driven by propaganda and disinformation and is destructive to welfare. He said, “the government can at any time access private data or government agency data on grounds of sovereignty or public order. This has dangerous implications.” Further, Jaiveer Shergill, a renowned Supreme Court Lawyer, also expressed his concerns about the gaps in this version of the Personal Data Protection Bill.
Following the previous draft’s criticism and the need for a more comprehensive version, the Bill was withdrawn on August 3, 2022.
The government unveiled the first draft of the revised Digital Personal Data Protection (PDP) Bill 2022 on November 18. The 2022 bill will focus on predefined principles that form the foundation of personal data protection and will inculcate a better understanding of data protection among the public and businesses. Fundamentally, the bill bifurcates the two sides involved in any situation related to processing digital personal data as a data fiduciary and a data principal.
The obligations of a Data Fiduciary
The Personal Data Protection Bill describes a “Data Fiduciary” as a person (or group of people) who determine the purpose and the means of processing data. There are several general obligations that a data fiduciary must follow while processing any personal data. First and foremost, a data fiduciary is responsible for complying with this Bill’s provisions while processing data or hiring a data processor on its behalf.
Secondly, a data fiduciary shall ensure that the processed information is “accurate and complete,’ if it is to be utilized for decision-making or passed on to someone else. He/she/they shall protect the data under possession and notify the regulatory Board in case of a data breach. Lastly, the fiduciary can only process data if the data principal has given consent.
The Personal Data Protection Bill describes a “Data Principal” as an individual whose personal data is to be processed. In the case of a child, the parents become the principal. The Bill provisions several rights and responsibilities that a data principal must have. Some of these rights include:
Right to Information about Personal data: As per this provision, the data principal shall have the right to obtain confirmation whether the fiduciary is processing or has processed the data. They can also request a summary of the processed data.
Right to correction and erasure of personal data: Under this provision, a principal shall request correction and erasure of their personal data while complying with the applicable laws. In such a request, the fiduciary is obliged to correct/complete/update/erase the requested data. Moreover, the data principal also has the right to grievance redressal if unsatisfied with the fiduciary’s responses.
While data principals have many protective rights, they must also abide by some duties. A data principal shall comply with all the Personal Data Protection Bill provisions and under no circumstances furnish any false or misleading information or impersonate another person. All the necessary information should also be verifiably authentic; only then can the data principal exercise their rights to correction or erasure under this Bill.
The 2022 draft will comply with a Board called the Data Protection Board of India to be established by the Central Government. The Board shall allocate work, receive grievances, and pronounce decisions digitally. It is the Board’s responsibility to ensure compliance with all provisions of this Act, act on valid principal grievances, and manage fiduciaries.
The latest Personal Data Protection Bill has successfully reworked some prior legislation gaps by incorporating hefty penalties for non-compliance and has relaxed specific rules on cross-border data flows that could be significant for large-scale enterprises. However, there are a few potential red flags. The Bill provides a near-exemption blanket for government-owned agencies from complying with some of the requirements. Additionally, the Bill suggests a dilution of the proposed Data Protection Board.
MeitY is optimistic that the latest Personal Data Protection Bill strikes a balance between fiduciaries and principals while staying aligned with the Supreme Court’s ruling on privacy and related fundamental human rights. The Ministry also invites public feedback on the draft until December 17, 2022.
Meta has announced it is supporting the Federation of Indian Chambers of Commerce & Industry (FICCI) with a sum of $1 million for the XR (extended reality) Open Source (XROS) fellowship program. The fellowship program will help 100 Indian developers work on extended reality technologies for the future of the metaverse.
Established by the Ministry of Electronics & Information Technology (MeitY), the National e-Governance Division will act as the technical partner for the program.
“The metaverse will not be built by one company alone. Through programs like XR Open Source, we will support Indian developers working on these exciting technologies,” said Nick Clegg, President of Global Affairs, Meta.
The #metaverse won’t be built by one company alone. Through programs like XR Open Source, we will support Indian developers working on these exciting technologies: Mr @nickclegg, President, Global Affairs, @Meta at FICCI XROS Fellowship Program supported by Meta & @NeGD_GoI. pic.twitter.com/t1ZzKSxNdI
The program will lay the foundation for India-specific solutions that are localized, affordable, and appropriate to regional languages.
The XR Open Source program is the second one in India through which Meta wants to boost the developer ecosystem pertaining to immersive technologies and will further the open ecosystem to create the metaverse.
The XROS program is part of Meta’s Research Fund and global XR Programmes, under which the company declared a $2 million fund along with MeitY Startup Hub earlier this year.
