Last Week in AI #160: Super fast 3D perception from Nvidia, Ukraine uses face recognition to identify dead Russian soldiers, US-China AI collaboration drops, and more!
Last Week in AI #160: Super fast 3D perception from Nvidia, Ukraine uses face recognition to identify dead Russian soldiers, US-China AI collaboration drops, and more!
Nvidia improves performance of NeRF by many orders of magnitudes, Ukraine uses facial recognition to identify dead Russian soldiers and notifies their families, dip in US-China co-authored AI papers
Last Week in AI #160: Super fast 3D perception from Nvidia, Ukraine uses face recognition to identify dead Russian soldiers, US-China AI collaboration drops, and more!
Top News
Nvidia shows off AI model that turns a few dozen snapshots into a 3D-rendered scene
Neural Radiance Fields (NeRF) is a technique for representing 3D scenes with neural networks. The NeRF neural network is trained with camera images of a scene, and it can then be used to extract information about the scene, like geometry, materials, and lighting, as well as render views of the scene from new camera angles. Ever since its inception in 2020, NeRF has taken the world of computer vision by storm, with a myriad of papers improving its performance and exploring creative applications. This new paper from Nvidia resolves one of NeRF’s most frustrating drawbacks, which is its slow training and rendering time. Combining a smaller NeRF network with a hash table, Nvidia’s researchers reduced the NeRF training time from hours to seconds, and rendering time from minutes to tens of milliseconds.
Our take: Achieving these many order of magnitudes speed-ups is an important advance in the “application” phase of NeRF that will see its deployment in real applications outside of research labs. It’s very likely that within another year or two, we will see NeRF used in a wide range of actual products in areas like 3D content creation, video games, VR, building inspections, video conferencing, robotics, and more.
Ukraine uses facial recognition to identify dead Russian soldiers, minister says
Reuters exclusively reported that Ukraine's Ministry of Defense this month began using technology from Clearview AI, a New York-based facial recognition provider that finds images on the web that match faces from uploaded photos. Now, Mykhailo Fedorov, Ukraine's vice prime minister who also runs the ministry of digital transformation, told Reuters Ukraine had been using Clearview AI software to find the social media accounts of dead Russian soldiers. From there, authorities are messaging relatives to make arrangements to collect the body, he said.
"As a courtesy to the mothers of those soldiers, we are disseminating this information over social media to at least let families know that they they've lost their sons and to then enable them to come to collect their bodies," Fedorov said in an interview, speaking via a translator.
Our take: Use of Clearview AI’s technology has been steadily increasing among US law enforcement, but this is the first reported case in which it has been used in a war. As facial recognition technology of this quality is relatively new, it is likely that this instance also represents one of the first cases of using AI to identify KIA soldiers. The fact that Ukraine can use Clearview AI’s services to identify Russian soldiers, despite the company being US-based, showcases just how far reaching their database is. While this is clearly a positive way to use the technology, it also highlights once again how far reaching the implications of such technology can be — especially when it is not constrained by laws and regulations.
US-China collaboration in AI papers drops amid ongoing tech war, Stanford report shows
Souring US-China relations have begun to manifest in their collaborations on AI papers. According to a report from Stanford's Institute for Human-Centered Artificial Intelligence, 2021 saw 9,660 papers co-authored by researchers from US and Chinese institutions. While this number is a decrease from the 10,000 papers US and Chinese researchers co-wrote the year before, US-China AI cooperation still leads the world. Additionally, while China leads the US in its number of AI publications and patent applications, the US remains ahead in citations and granted patents
Our Take: It's expected that worsening US-China relations and the resulting decouplings would show up in numbers describing their cooperation. The fact that the US and China still cooperate in AI research more than any other pair does speak to the realities of their integration. It seems unlikely that the two nations will be able to entirely decouple, economically or in research. But it will be interesting to watch how much AI research begins to diverge between the two countries and how much this tracks the status of their relationship.
Other News
Research
Google AI and UC Berkely Researchers Introduce A Deep Learning Approach Called 'PRIME' That Generates AI Chip Architectures By Drawing From Existing Blueprints And Performance Figures - "This data-driven optimization approach generates AI chip architectures by using logged data without further hardware simulation."
Researchers From Cortical Labs Develop DishBrain: A Neural Network With Biological Neurons - "Researchers at Cortical Labs, Australia propose that integrating neurons into digital systems to tap on their inherent intelligence could enable performance that would be impossible to achieve with silicon alone and provide insight into the biological origins of intelligence."
Seeing an elusive magnetic effect through the lens of machine learning - "An MIT team incorporates AI to facilitate the detection of an intriguing materials phenomenon that can lead to electronics without energy dissipation."
ML Researchers From Oxford Propose a Forward Mode Method to Compute Gradients Without Backpropagation - "Oxford University researchers have proposed a novel method that might cut training time in half."
Machine Learning Can Predict Wind Energy Efficiency - "In an article recently published in the journal Energies, researchers presented a comparative study of efficient wind power prediction using machine learning methods. "
Machine Learning Techniques Can Speed Up Glacier Modeling By A Thousand Times - "Since meltwater from glaciers and ice sheets is a major component of sea level rise, models like this are a valuable tool to assess their potential future contribution."
AI confirms the obvious: The pandemic bummed people out - “Scientists used machine learning to make a mood ring for the internet, and it showed us how sad the world was at the start of the pandemic.”
Applications
Ubisoft shows off machine learning tool for modelling animals - "Ubisoft has recently shown off its ZooBuilder AI tool, a piece of software that can take footage of animals and make skeletons for in-game models. "
Humility over hype: An AI leader at Google lays out a vision to transform medical care - "From a high perch within Google, Greg Corrado is steering the development of artificial intelligence tools he believes will dramatically improve health care in coming years."
Robotic exoskeleton uses machine learning to help users with mobility impairments - "An important element of the new device is technology that allows the skeleton to effectively guess the intentions of the user."
Kroger and Nvidia partner to ‘reinvent the shopping experience’ with AI and digital twin simulations - "Nvidia and Kroger today announced a “strategic collaboration” designed to bring more AI-powered applications and services to the grocery realm."
Business
Mercedes Drive Pilot Beats Tesla Autopilot By Taking Legal Responsibility - "Mercedes will accept full legal responsibility for the vehicle whenever Drive Pilot is active. The automaker hopes to offer the system in the U.S. by the end of 2022."
Meet the DeepMind mafia: These 17 alumni from Google's AI research lab are raising millions for their own startups, from climate to crypto - "Over the past decade, Google-owned artificial intelligence research lab DeepMind has become a repository for top AI talent."
Israel's OurCrowd set to open AI R&D centre in Abu Dhabi - "OurCrowd, one of Israel's largest venture firms, plans to open an artificial intelligence-based research and development centre in Abu Dhabi by early June that will help with investment decision making, a company official said."
Apple Is Among Most-Trusted Brands to Develop Autonomous Vehicles - "Despite not having a vehicle of its own on sale, Apple landed in the middle of the top ten with 13 percent of consumer's trust."
A.I. creative platform D-ID, the tech behind those viral videos of animated family photos, raises $25M - "D-ID, the Israeli company leveraging A.I. to create unique, viral experiences like “Deep Nostalgia,” which animates the faces of long-lost relatives in your old photos, announced today the raise of a $25 million Series B round of funding led by Macquarie Capital."
Nvidia unveils new technology to speed up AI, launches new supercomputer - "Nvidia announced new chips and technologies that it said will boost the computing speed of increasingly complicated artificial intelligence algorithms, stepping up competition against rival chipmakers vying for lucrative data center business."
Top AI execs, including Richard Socher, launch AIX Ventures - "In a sign that the coffers of AI startups need little replenishing, Richard Socher, former chief scientist at Salesforce and the CEO of You.com, today announced the launch of a $50 million “AI-focused” venture fund called AIX Ventures."
Mayo launches AI startup program, with assists from Epic and Google - "The Mayo Clinic Platform_Accelerate is a 20-week initiative aimed at helping four companies in its initial cohort become market-ready."
Concerns
How Native Americans Are Trying to Debug A.I.’s Biases - "Data on Native communities are not at the levels needed for accuracy in A.I.-driven tools. A group is trying to solve that problem. "
Software vendors are pushing "explainable A.I." that often isn't - "In some cases, explanations are being used to engender trust in A.I. when it isn't warranted, researchers say."
Banks Using AI Are Ripe for Russian Sabotage - "Banks and other financial institutions utilizing artificial intelligence may be uniquely susceptible to retaliatory Russian cyberattacks as taxing international sanctions worsen, experts warn."
Automation will erase 'knowledge jobs' before most blue collar jobs: Future Today Institute CEO - "Contrary to popular belief, high-skill professions may be more susceptible to replacement by artificial intelligence, according to Future Today Institute CEO Amy Webb."
Analysis
12 Graphs That Explain the State of AI in 2022 - "The 2022 AI Index talks jobs, investments, ethics, and more"
How self-driving cars got stuck in the slow lane - "The technology behind autonomous vehicles has proved devilishly hard to perfect. And progress hasn’t been helped by Tesla boss Elon Musk’s army of superfans"
Examining the ethics of AI development - "AI technologies will undoubtedly influence and impact our lives, beliefs and culture. The role of AI ethicists should be to ensure that these technologies wield that impact in an equitable and benevolent way."
Invading Ukraine has upended Russia's A.I. ambitions—and not even China may be able to help - "Russia President Vladimir Putin’s invasion of Urkaine will likely spell disaster for the country’s ambitions to be a leader in artificial intelligence."
Report: 70% of U.S. consumers want to use AI for their jobs - "In nine categories of work tasks, more workers wanted AI to help than wanted AI to do everything or do nothing. "
Expert Opinions
Meta’s Yann LeCun strives for human-level AI - "In a recent event held by Meta AI, LeCun discussed possible paths toward human-level AI, challenges that remain and the impact of advances in AI."
Deep Learning on Electronic Medical Records is doomed to fail - "A few years ago, I worked on a project to investigate the potential of machine learning to transform healthcare through modeling electronic medical records. I walked away deeply disillusioned with the whole field and I really don’t think that the field needs machine learning right now."
Andrew Ng predicts the next 10 years in AI - "We connected with Ng to discuss what he calls the “data-centric approach” to AI, and how it relates to his work with Landing AI and the big picture of AI today."
Copyright © 2022 Skynet Today, All rights reserved.