Gifu City, a demonstration experiment for cashless payments on an autonomous driving bus by face authentication

　Gifu City has been conducting a civilian participation -type demonstration experiment for the third consecutive year since 2019, aiming to introduce automatic driving technology to public transportation.As a background, the number of elderly people who do not have a driver's license, the increase in traffic vulnerable people who are difficult to move to the bus stop, and the difficulty of sustainability of regional traffic due to a shortage of bus drivers.

　In FY2011, the company adopted a French vehicle without driver's seat, handle, accelerator, and brake pedal, with the goal of reducing the opportunity for drivers to reduce the number of opportunities for drivers rather than experimenting in 2020.

　Next, at the same time as the driving verification of the autonomous driving bus, a demonstration experiment of cashless payments using face authentication AI will be conducted.The facial authentication AI engine adopted the triple eyes image recognition base "AIZE" (Photo 1).

　The demonstration experiment will be conducted on 9 days from October 23.Run on a general road with people on a small bath type electric vehicle.The route departs from Gifu City Hall, returns to Gifu City Hall on Gifu City Hall on Nagara Bridge Street via Kinka Bridge Street via JR Gifu Station, departs from Gifu City Hall and returns to Gifu City Hall via Wakamiya Street.I prepared two patterns of 2 kg.

　At that time, verify that the traffic light color is automatically judged by the communication between the vehicle and the traffic light, and that the carrier and the obstacle can be automatically detected from the vehicle -mounted sensor.The running speed is assumed to be up to 19 km / h.The driver can also use the controller to operate the vehicle.

　Participants in the demonstration experiment will shoot face images with a camera such as iPad in advance and register with AIZE.When the face is authenticated with a tablet installed on the bus, the captured image is analyzed with AIZE on the cloud and confirm the person.In the case of a registered person in advance, the fare is displayed on the screen along with the message "Thank you for boarding".This allows you to confirm that cashless payment has been performed.At the same time as the face authentication, the temperature will be performed, supporting infectious diseases in the bus.

　In the future, Gifu City has been conducting experiments for practical use, such as remote monitoring of the inside of the car and face authentication with a view to cashless payments, assuming the operation of a "bus without crew"."As the population declines and the declining birthrate and aging population, the city aims to create a compact network that makes citizens easy to live in with networks such as highly convenient public transportation in the central city and regions."

　AIZE is an image recognition system by deep learning (deep learning) born from the research of Go AI worked on Triple Eyes.Send image data to the cloud and analyze.512 Democraticity of dimensions is detected from the face image and individual recognition is recognized.The person is authenticated with a probability of 98 % or more when wearing a mask, and the age, gender, and emotions can be recognized.