A research team led by Professors Ki-Hun Jeong and Doheon Lee from the Korea Advanced Institute of Science and Technology (KAIST) Department of Bio and Brain Engineering reported the development of a technique for facial expression detection by mixing light-field camera techniques with artificial intelligence (AI) technology.

Unlike a conventional camera, the light-field camera contains sets of micro-lens (微透镜) in front of the image sensor, which makes the camera small enough to fit into a smart phone while allowing it to acquire the spatial and directional information of the light with a single shot. The technique has received attention as it can reconstruct images in a variety of ways including multi-views, refocusing and 3D image acquisition, giving rise to many potential applications.

However, the optical (光学的) crosstalk between shadows caused by external light sources in the environment and the micro-lens has limited existing light-field cameras from being able to provide accurate image contrast and 3D reconstruction.

The research team applied a laser in the near-infrared (NIR) range to stabilize the accuracy of 3D image reconstruction that previously depended on environmental light. When an external light source is shone on a face at 0-30, and 60-degree angles, the light-field camera reduces 54% of the image reconstruction errors. Additionally, by inserting a light-absorbing layer for visible and NIR wavelengths between the sets of micro-lens, the team could minimize optical crosstalk while increasing the image contrast by 2.1 times.

Through this technique, the team could overcome the limitations of existing light-field cameras and develop a more advanced NIR-based light-field camera (NIR-LFC) improved for the 3D image reconstruction of facial expressions. Using the NIR-LFC, the team acquired high-quality 3D reconstruction images of facial expressions expressing various emotions regardless of the lighting conditions of the surrounding environment.

The facial expressions in the acquired 3D images were distinguished through machine learning with an average of 85% accuracy-a statistically significant figure compared to when 2D images were used. Furthermore, by calculating the inter-dependency of distance information that varies with facial expression in 3D images, the team could identify the information alight-field camera uses to distinguish human expressions.

Visitors to HENN-NA, a restaurant outside Nagasaki, Japan, are greeted by an odd sight:   their food being prepared by a row of humanoid robots that look like the Terminator. H. I. S., the company that runs the restaurant, as well as a nearby hotel where robots check guests into their rooms and help with their luggage, turned to automation partly out of necessity. Japan’s population is shrinking, and its economy is booming; the unemployment rate is only 2.8 percent. “Using robots makes a lot of sense in a country like Japan,” said CEO Hideo Sawada.

Sawada predicts that 70 percent of the jobs at Japan’s hotels will be automated in the next five years. “It takes about a year to two to get your money back,” he said. “But since you can work them 24 hours a day, and they don’t need vacation, eventually it’s more cost-efficient to use the robot.”

This may seem like a vision of the future best suited—perhaps only suited—to Japan. But according to Michael Chui, a partner at the McKinsey Global Institute, many tasks in the food-service and accommodation industry are exactly the kind that are easily automated. Chui’s latest research estimates that 54 percent of the tasks workers perform in American restaurants and hotels could be automated using currently available technologies.

The robots, in fact, are already here. Chowbotics, a company in Redwood City, California, manufactures Sally, a boxy robot that prepares salads ordered on a touch screen. Botlr, a robot butler, now brings guests extra towels and toiletries in dozens of hotels around the country.

This seems to be worrying. America’s economy isn’t developing nearly as smoothly as Japan’s, and one of the few bright spots in recent years has been employment in restaurants and hotels, which have added more jobs than almost any other industry. That growth, in fact, has helped dull the blow that automation has delivered to other industries. The food-service and accommodation industry now employs 13.7 million American. Since 2013, it has accounted for more jobs than manufacturing.

These new positions once seemed safe from robots because they required a human touch in a way that manufacturing or mining jobs did not. When ordering a coffee or checking into a hotel, human beings want to interact with other human beings — or so we thought. The companies bringing robots into the service industry are betting that we’ll be happy to trade our relationship with human waiters or clerks for greater efficiency. They’re also confident that adding robots won’t necessarily mean cutting human jobs.

After COVID-19, the Bureau of Labor Statistics predicts that while the number of nurses has increased in the past three years, there is still a shortage of registered nurses, and that there will be over one million unfilled nursing jobs. So what’s the solution? Robots.

Japan is ahead of the curve when it comes to this trend. Toyohashi University of Technology has developed Terapio, a robotic medical cart that can make hospital rounds, and deliver medicines and other items . This type of robot will likely be one of the first to be used in hospital.

Robots capable of social engagement help with loneliness as well as cognitive (认知) functioning. Telepresence robots such as MantaroBot, Vgo, and Giraff can be controlled through a computer, smartphone, or tablet, allowing family members or doctors to remotely monitor patients or Skype (网络电话) them. If you can’t get to the nursing home to visit grandma, you can use a telepresence robot to hang out with her.

A robot’s appearance affects its ability to successfully interact with humans, which is why the Human-Interactive Robot Research decided to develop a robotic nurse that looks like a huge teddy bear. RIBA, also known as “Robear, ” can help patients into and out of wheelchairs and beds with its strong arms.

On the less cute and more scary side there is Actroid F, which has such soft skin and natural hair color that some patients may not know the difference. This conversational robot companion has cameras in its eyes, which allow it to track patients and use appropriate facial expressions and body language in its interactions.

It’s important to note that robotic nurses don’t decide courses of treatment or make diagnoses. Instead, they perform routine and laborious tasks, freeing nurses up to attend to patients with immediate needs. This is one industry where it seems the integration of robots will lead to cooperation, not replacement.