Google Maps will use AI to update store hours
14/04/2022Google Maps will update the hours of more than 20 million businesses around the world in the next six months. To do this, they will use Artificial Intelligence (AI) and a self-developed machine learning model, which identifies shop hours that are not updated. Through this system, the AI will update the hours according to itsel predictions based on parameters such the last update of the shop's profile, the hours of nearby stores and the busiest hours in that location. In addition, the algorithm will take datas from Google Business, information from the store's website, if it has one, and images from Street View. In some countries, Google will also use Google Duplex conversational AI technology to make calls with the quality of a human voice to the stores and confirm their opening hours. In this list of countries, France will be the only European country where the telephone verification will be carried out. The rest of the countries are Argentina, Australia, Chile, Japan, Mexico, New Zealand, Peru and the United States.
[ ... ]
New Wifi technology to accurately count seated people
27/09/2021Small natural gestures such crossing your legs, checking your mobile, or scratching your nose are enough to count the number of people sitting in certain space. A group of researchers from the University of California have developed a system that overcomes the precision problems that the lack of movement during most of the time in seated people caused in this type of counting operations. This new system also allows those seated behind a wall to be counted. The new technology that is suitable for Wi-Fi devices is based on a mathematical model that, in a similar way to how the queue management theory works, relates the total number of gestures to the number of people sitting in the same space. Accuracy is 96.3, based on tests performed during the study. In addition, the researchers say, the system offers an accuracy of almost 90%, in the detection of people behind the walls or dividing columns of any space.
[ ... ]