Welcome to the fourth edition of the AI Index Report. This year we significantly expanded the amount of data available in the report, worked with a broader set of external organizations to calibrate our data, and deepened our connections with the Stanford Institute for Human-Centered Artificial Intelligence (HAI). The AI Index Report tracks, collates, distills, and visualizes data related to artificial intelligence. Its mission is to provide unbiased, rigorously vetted, and globally sourced data for policymakers, researchers, executives, journalists, and the general public to develop intuitions about the complex field of AI. The report aims to be the most credible and authoritative source for data and insights about AI in the world.

Deep object recognition models have been very successful over benchmark datasets such as ImageNet. How accurate and robust are they to distribution shifts arising from natural and synthetic variations in datasets? Prior research on this problem has primarily focused on ImageNet variations (e.g., ImageNetV2, ImageNet-A). To avoid potential inherited biases in these studies, we take a different approach. Specifically, we reanalyze the ObjectNet dataset recently proposed by Barbu et al. containing objects in daily life situations. They showed a dramatic performance drop of the state of the art object recognition models on this dataset. Due to the importance and implications of their results regarding the generalization ability of deep models, we take a second look at their analysis. We find that applying deep models to the isolated objects, rather than the entire scene as is done in the original paper, results in around 20-30% performance improvement. Relative to the numbers reported in Barbu et al., around 10-15% of the performance loss is recovered, without any test time data augmentation. Despite this gain, however, we conclude that deep models still suffer drastically on the ObjectNet dataset. We also investigate the robustness of models against synthetic image perturbations such as geometric transformations (e.g., scale, rotation, translation), natural image distortions (e.g., impulse noise, blur) as well as adversarial attacks (e.g., FGSM and PGD-5). Our results indicate that limiting the object area as much as possible (i.e., from the entire image to the bounding box to the segmentation mask) leads to consistent improvement in accuracy and robustness.

Recently, I was at a party in San Francisco when a man approached me and introduced himself as the founder of a small artificial intelligence (AI) start-up. As soon as the founder figured out that I was a technology writer for The New York Times, he launched into a pitch for his company, which he said was trying to revolutionise the manufacturing sector using a new AI technique called "deep reinforcement learning". The founder explained that his company's AI could run millions of virtual simulations for any given factory, eventually arriving at the exact sequence of processes that would allow it to produce goods most efficiently. This AI, he said, would allow factories to replace entire teams of human production planners, along with most of the outdated software those people relied on. "We call it the Boomer Remover," he said.

I wake up in the middle of the night. "Hey, Google, what's the temperature in Zone 2," I say into the darkness. A disembodied voice responds: "The temperature in Zone 2 is 52 degrees." "Set the heat to 68," I say, and then I ask the gods of artificial intelligence to turn on the light. Many of us already live with A.I., an array of unseen algorithms that control our Internet-connected devices, from smartphones to security cameras and cars that heat the seats before you've even stepped out of the house on a frigid morning.

Artificial intelligence is the technological blow that took the world by storm. When the term'artificial intelligence' was first coined at a conference, …

Currency notes have identifiers that allow the visually impaired to identify them easily. This is a learned skill. On the other hand, classifying them using images is an easier solution to help the visually impaired identify the currency they are dealing with. Here, we use pictures of different versions of the currency notes taken from different angles, with different backgrounds and covering different proportions. The dataset contains 195 images of 7 categories of Indian Currency Notes -- Tennote, Fiftynote, Twentynote, 2Thousandnote, 2Hundrednote, Hundrednote, 1Hundrednote.

I have been reproducing (or trying to reproduce) times series classification results for 23 years. In general, less than half the papers can be reproduced, but since the advent of deep learning, the fraction has gotten worse. A) "We made a best faith effort to reproduce the results in [x], but were unable to do so, thus we omit it from comparison" B) "We made a best faith effort to reproduce the results in [x], but were unable to do so. We do place our results using this method in Table Y with an asterisk, to denote that this is our best understanding of the algorithms performance, but just not reflect the claimed accuracy possible in [x]".

When Krish Ashok, a technologist and food science enthusiast based out of Chennai, shared pictures of his father and great-grandmother, animated through a new artificial intelligence (AI) tool with the senior members in his family, they were deeply moved. "They were shaken by what they saw," Ashok told Zenger News. Deep Nostalgia, which animates old photos, is the latest AI tool that has taken the Internet by storm. But, it has also sparked a conversation about the ethical implications of such tools -- and not everyone is comfortable with it. The video reenactment technology's web-based version was overrun by eager users within a week of its launch on Feb. 25, and the site had to undergo maintenance to keep up.

In the Zagar & Evans pop song classic In the Year 2525, the duo sang of a year where "some machine is doing that for you." A company called Sonantic has created what it claims is the first artificial intelligence voice models that sound genuinely human and capable of expressing "a wide range of complex human emotions, from fear and sadness to joy and surprise." Gaming producers like Obsidian Entertainment, Splash Damage and 4A Games are already on board with the technology, which is in use from development through post-production. The concept is similar to CGI for audio. Human, realistic voice tech is in its infancy, but along with this brave new frontier is a new opportunity for voiceover artists – banking their words for future use.

And I also think there's going to be a different conversation about AI, machine learning and things like that.