There are a number of uses for knowing what, where and how many images of your company’s logo are out on the internet. Until now the ability to know this information has been poor, and detection systems not of the greatest quality.
Recent research has been carried out and has tested the different methods of detection and the results of each, and also the results when using combined detection systems. The results have been published and as the methods get absorbed into the software and search engines, there will be a significant change in the data available to businesses.
Searches for the logo could be used to identify where a logo shouldn’t be being used, or to see what the penetration of a logo is in a certain area, eg. country domain etc.
The research by Hang Su, Shaogang Gong, Xiatian Zhu into Scalable Deep Learning Logo Detection brought the comparably low (SL2 W/O CE) 28.9% to 46.9% (Co Learning (YOLO)(SL2) The Co learning Faster R-CNN method coming in at a close 44.2%
So if you’re looking to tell what impact your new logo has reached in the cyberspace, that answer may soon be at the tip of your fingers.