Pre-training on massive video datasets has become essential to achieve high action recognition performance on smaller downstream datasets. However, most large-scale video datasets contain images of people and hence are accompanied with issues related to privacy, ethics, and data protection, often preventing them from being publicly shared for reproducible research. Existing work has attempted to alleviate these problems by blurring faces, downsampling videos, or training on synthetic data. On the other hand, analysis on the {\em transferability} of privacy-preserving pre-trained models to downstream tasks has been limited. In this work, we study this problem by first asking the question: can we pre-train models for human action recognition with data that does not include real humans? To this end, we present, for the first time, a benchmark that leverages real-world videos with {\em humans removed} and synthetic data containing virtual humans to pre-train a model. We then evaluate the transferability of the representation learned on this data to a diverse set of downstream action recognition benchmarks. Furthermore, we propose a novel pre-training strategy, called Privacy-Preserving MAE-Align, to effectively combine synthetic data and human-removed real data. Our approach outperforms previous baselines by up to 5\% and closes the performance gap between human and no-human action recognition representations on downstream tasks, for both linear probing and fine-tuning.

Pre-training on massive video datasets has become essential to achieve high action recognition performance on smaller downstream datasets. However, most large-scale video datasets contain images of people and hence are accompanied with issues related to privacy, ethics, and data protection, often preventing them from being publicly shared for reproducible research. Existing work has attempted to alleviate these problems by blurring faces, downsampling videos, or training on synthetic data. On the other hand, analysis on the {\em transferability} of privacy-preserving pre-trained models to downstream tasks has been limited. In this work, we study this problem by first asking the question: can we pre-train models for human action recognition with data that does not include real humans?

Led by Joon Sung Park, a Stanford PhD student in computer science, the team recruited 1,000 people who varied by age, gender, race, region, education, and political ideology. They were paid up to 100 for their participation. From interviews with them, the team created agent replicas of those individuals. As a test of how well the agents mimicked their human counterparts, participants did a series of personality tests, social surveys, and logic games, twice each, two weeks apart; then the agents completed the same exercises. The results were 85% similar.

Design of Artificial Intelligence and robotics habitually assumes that adding more humanlike features improves the user experience, mainly kept in check by suspicion of uncanny effects. Three strands of theorizing are brought together for the first time and empirically put to the test: Media Equation (and in its wake, Computers Are Social Actors), Uncanny Valley theory, and as an extreme of human-likeness assumptions, the Singularity. We measured the user experience of real-life visitors of a number of seminars who were checked in either by Smart Dynamics' Iwaa, Hanson's Sophia robot, Sophia's on-screen avatar, or a human assistant. Results showed that human-likeness was not in appearance or behavior but in attributed qualities of being alive. Media Equation, Singularity, and Uncanny hypotheses were not confirmed. We discuss the imprecision in theorizing about human-likeness and rather opt for machines that 'function adequately.'

Blink-182's Travis Barker is widely regarded as one of the best drummers in the music industry. But he could face some stiff competition from a rather unexpected musician - Xiaomi's humanoid robot, CyberOne. A new video shows the $104,000 robot drumming along to a song with expert precision. Drumming isn't CyberOne's only skill - it also has arms and legs that allow it to walk just like a real human, while the android is fitted with AI technology that allows it to detect 45 human emotions. A new video shows Xiaomi's humanoid robot, CyberOne drumming along to a song with expert precision In the short clip, which was posted to YouTube by IEEE Spectrum, CyberOne can be seen sitting behind a drum kit, playing along to a pre-recorded track.

For the past two years, I've been following a woman around the internet. It sounds ominous, I know, but hear me out. Her name is Albertina Geller, and I first stumbled across her online in October 2020, on LinkedIn. She'd listed herself as a "self-employed freelancer" in Chicago. I'm also a self-employed freelancer, so we had that in common. In her bio, she said that "I learn & teach people how to be healthy, balance their gut and improve their immune system for healthy living." I've had some gut and immune-system issues myself. It was a connection practically written in the stars. But I have to admit that what first interested me about her -- what led me to spend two years tracking her, at a distance -- wasn't our shared interests. Her LinkedIn photo was a straight-on headshot of a white woman, mid- to late 20s, with a pale complexion and lightly rosy cheeks. She had shoulder-length blond hair, swept neatly to one side.

From'Ex Machina' to'I, Robot', humanoid robots have been staple features of science fiction blockbusters for years. Now, lifelike robots are becoming more and more popular in the real world, and the latest offering is one of the most impressive yet. Xiaomi has revealed its first humanoid robot – a $104,000 bot named CyberOne. CyberOne has arms and legs that allow it to walk just like a real human, while the android is fitted with AI technology that allows it to detect 45 human emotions. Its unveiling comes just one month before the highly anticipated launch of Tesla's humanoid robot, Optimus.

"Rising demand is driving the boom of digital humans," says Shiyan Li, head of the digital human and robotics business at Baidu, which created the digital model-actor, Gong. "In China alone, there are over 400 million ACGN (animation, comics, games, and novel) fans, and an enterprise market worth hundreds of billions of dollars centered on digital humans." And according to a company that tracks business registrations, Qichacha, China now has more than 280,000 enterprises that engage in digital human-related activities. The debut of Baidu's digital celebrity may not seem like much at first, as the concept of "virtual idols" has been around for years. For example, US virtual influencer Lil Miquela has been appearing alongside real human celebrities in online advertisements and TV commercials since 2016, gaining over three million Instagram followers.

This was the year we all began social distancing. But the ensuing isolation was already the norm for a rapidly growing population – and a major opportunity for many businesses. And as isolation has engulfed the globe like the virus itself, the business of loneliness is booming. Even before the pandemic, loneliness had been deemed an official epidemic in several countries. Rates of loneliness in the US have doubled over the past 50 years.

In this paper, we introduce Watch-And-Help (WAH), a challenge for testing social intelligence in agents. In WAH, an AI agent needs to help a humanlike agent perform a complex household task efficiently. To succeed, the AI agent needs to i) understand the underlying goal of the task by watching a single demonstration of the humanlike agent performing the same task (social perception), and ii) coordinate with the humanlike agent to solve the task in an unseen environment as fast as possible (human-AI collaboration). For this challenge, we build VirtualHome-Social, a multi-agent household environment, and provide a benchmark including both planning and learning based baselines. We evaluate the performance of AI agents with the humanlike agent as well as with real humans using objective metrics and subjective user ratings. Experimental results demonstrate that the proposed challenge and virtual environment enable a systematic evaluation on the important aspects of machine social intelligence at scale. Without much prior experience, children can robustly recognize goals of other people by simply watching them act in an environment, and are able to come up with plans to help them, even in novel scenarios. In contrast, the most advanced AI systems to date still struggle with such basic social skills. In order to achieve the level of social intelligence required to effectively help humans, an AI agent should acquire two key abilities: i) social perception, i.e., the ability to understand human behavior, and ii) collaborative planning, i.e., the ability to reason about the physical environment and plan its actions to coordinate with humans. In this paper, we are interested in developing AI agents with these two abilities. Towards this goal, we introduce a new AI challenge, Watch-And-Help (WAH), which focuses on social perception and human-AI collaboration. In this challenge, an AI agent needs to collaborate with a humanlike agent to enable it to achieve the goal faster. In particular, we present a 2-stage framework as shown in Figure 1. In the first, Watch stage, an AI agent (Bob) watches a humanlike agent (Alice) performing a task once and infers Alice's goal from her actions.