I-AUC has developed a track search that allows you to search by body type using image recognition AI.
Image recognition AI instantly judges the exhibited vehicle image and classifies it into 10 types of body types. It is possible to distinguish the body type even in the image of the exhibited vehicle that looks the same at first glance like a wing, van or refrigerated frozen vehicle.
In addition, body type search can be used not only at auctions specializing in trucks, but also at all venues connected to I-AUC.
The body type recognition rate of image recognition AI is as high as 95%, and most of the misidentifications are due to background images and the like, and the recognition rate has been improved to the same level as human judgment.
Mr. Takuyuki Kuroyanagi, CEO of Aucnet IBS, who worked on the development of image recognition AI, said, "I have been researching the possibility of image recognition from the truck shared inventory of group companies. With deep learning, AI will recognize the design and will be able to judge even if the content has different photo angles, "said the possibility of AI.
Yasuhiro Nakayama, General Manager of the Product Distribution Planning Office, said, "It is difficult to distinguish the body type of a truck from the model, and many members have requested" body type search. "I-AUC took this voice seriously and took this voice seriously, we started verification of image recognition AI to realize truck search optimization, and this time it was completed. We focused on making the body type search screen easy to understand. Making it easier to find the vehicle of the body type you are looking for. I think that it will lead to the reduction of opportunity loss for members. "
Mr. Takeshita, Dean of the Sales Department, said, "We have received feedback from members that truck search is good at the sales floor. We also hear joy from members when it is easy to search."
Mr. Nakayama said, "We will continue to make improvements every day, and while utilizing new tools such as AI tools, we will continue to make improvements that lead to" ease of searching "and" ease of buying. " I said.
Note: above described function is already implemented in "Ai-Net" by Arai Group.
Read more about auctions: https://www.akebono-akb.com/blog