Sony AI, the Japanese conglomerate’s machine learning and artificial intelligence (AI) research arm, has unveiled a new initiative to be dedicated to research and development in gastronomy. The flagship gastronomy project, as Sony AI has called it will concentrate on three main areas: an AI application for new recipe creation, a robotics solution that can support kitchen chefs, and an initiative for group co-creation. Sony AI said it would use data sources to create the recipe development app, which it hopes will be used by chefs when developing recipes and menus, including recipes and ingredient data, such as taste, scent, colour, molecular structure, and nutrients. The company added that it expects the chef assisting robot to be trained with sensors and AI so that it can assist chefs from planning to plating during the entire cooking process.
The community co-creation program, meanwhile will concentrate on collaborating with the worldwide chef community, as well as universities, research institutes, and relevant companies to further assist Sony AI in its project.
“What has become apparent with COVID-19 is the importance of sustainability and health, and the value of intangible cultural assets including gastronomy and the arts we wish to protect,” Sony AI CEO Hiroaki Kitano.
“Through the power of AI and robotics, we want to reaffirm the principle of our gastronomy flagship project, which is to enable creative gastronomy that is at the same time healthy and sustainable.”
In acknowledging that gastronomy is a new area for the company, Sony AI COO Hiroaki Kitano said there are plans to “pursue our activities in partnership with creator chefs, food experts, and researchers, as well as universities, research institutes, and companies with cutting-edge capabilities”.
Sony produced the first image sensor equipped with AI processing in the world earlier this year. The latest image sensor has a setup made up of a logic chip and a pixel chip. The logic chip, named Sony’s digital signal processor, is dedicated to the processing of AI signals, along with the AI model’s memory. Sony said that by using AI, the sensor can also output metadata — semantic information belonging to image data — instead of image information, which decreases the amount of data and fixes privacy issues. Sony said the logic chip inside the image sensor for realistic applications allows it to send data within a single video frame for each individual output image frame for AI processing.
In November of last year, Sony AI was formed as an internal organization within Sony Corporation before it was launched as a new company in April.