Artificial Intelligence to Assess the quality of Sushi

A Japanese chain of sushi restaurants is using an artificial intelligence-powered application. With it, they can assess the quality of tuna key steps in the preparation of Sushi. Traditionally Sushi requires years of trading from experienced human buyers.

The application’s name is Tuna Scope. The Japanese advertising firm Dentsu Inc developed it; the firm uses machine learning algorithms trained on thousands of pictures of the cross-section of tuna tails. These tails are cut off the meat, which may reveal much concerning a fish’s constitution.

The application grades the tuna on a five-point scale from a single picture. It makes the measurement based on visual characteristics like the layering of fat and the sheen of the flesh. For an experienced fish grader, those attributes speak volumes concerning the sort of life the fish led, how active it was, and what it ate – thus, they can find out the resulting flavor. Dentsu says that its artificial intelligence has captured the unexplainable nuances of craft tuna examinations. Moreover, comparing the application with human buyers, the application issued the same grade more than four times in five.

Nevertheless, fishmongers and experts are a little more cautious concerning Tuna Scope’s ability to replace fish grades. Critiques came primarily from those who buy meat for high-end sashimi and sushi.

Keiko Yamamoto is a sushi instructor and chef. He works in London. She told the Verge that it is undoubtedly possible to grade tuna based on visuals alone. Nevertheless, they often judge the quality of products based on touch as well. Moreover, Yamamoto says that the appearance of tuna is everything. She added that she has had to cut fresh tuna every two weeks. Thus, she knows what is not good and what is right.

Sushi

Yamamoto said that the exact qualities buyers look for is hard to capture in words. Nevertheless, they are unmistakable to the trained eye. The highest-quality tuna has a certain degree of transparency and intense bright red color, as if the flesh is almost glowing. It looks soft, or bouncy to your eye, says Yamamoto. Good quality tuna is shiny and silkier.

Thus, it seems possible to use artificial intelligence to make the necessary assessments of quality. Yamamoto is not surprised that Japan is pursuing that tech. Aging populations means traditional skills don’t always go down to younger generations.

According to The Asahi Shimbun, right now, it seems that Tuna Scope is only being used to grade fish for restaurant chain Kura Sushi. It uses other cost-saving devices like robotic dishwashers and offers cheap Sushi. Reportedly, Kura Sushi purchases seventy percent of its fish for Sushi overseas. However, it is wary concerning its buyers traveling during the current pandemic situation. The application means local agents can make on-the-spot assessments instead.

That sort of automation could work for a large chain like Kura Sushi. Nevertheless, it will not meet the demands of high-end chefs and sushi aficionados. This is according to Richard Cann.

He works at T&S Enterprises. Cann is a sales manager there. T&S Enterprises is a wholesale fishmonger that supplies many of the top Japanese restaurants in London.