An Evaluation of Reverse Image Search Performance of Google

Abstract

This study investigates reverse image search performance of Google, in terms of Average Precisions (APs) at various cut-off points, on finding out similar images by using fresh Image Queries (IQs) from the five categories Fashion, Computer, Home, Sports, and Toys, in order to have an insight about reverse image search performance of Google and then, motivate the researchers and inform the users. Five fresh IQs with different main concepts were created for each of the five categories. These 25 IQs were run on the search engine and for each, the first 100 images retrieved were evaluated with binary relevance judgment. APs at the cut-off points 20, 40, 60, 80, and 100 were calculated for each category and for all 25 IQs. The performance range is from similar to 42% for Toys category at the cut-off point 100 to 71% for Home category at the cut-off point 20. When the categories are ignored, Google's performance range is from similar to 52% at the cut-off point 100 to similar to 57% at the cut-off point 20. It seems that reverse image search performance of Google needs to be improved.