Google recently introduced query-by-example in their image search engine. Query-by-example allows users to drag an image from their desktop into the search input field and after hitting the “search” button a bunch of matching images is returned. I consider it a positive sign that visual retrieval methods from the academic literature are finally finding their way to real-world deployment at Google scale. For those interested, the first version of query-by-example was described in 1992 in a paper by Kato et al. and was made popular by IBM’s QBIC. See their demo at the website of the Hermitage. Query-by-example can be fruitful when users search for the same object under slightly varying circumstances and when the target images are available indeed.If proper example images are unavailable, query-by-example is not effective at all. Moreover, users often do not understand similarity of the low-level visual features used for recognition. They expect semantic similarity. I tried a search using a recent photo of me, Masoud and Stratis. Judging from the results I believe there is still plenty of research left to do before query-by-example is solved.