Image search is evolving rapidly. Today, machines understand much more about images than just a year ago: it can read the text on the image, see its colors and classify it based on its form, shape, and textures.
So which advanced image search methods can we use today?
Image Search Based on Image Content
Face search has been a hot topic recently.
- Exalead has become known for integrating facial recognition technology but it still lacks some accuracy. I tried searching [Paris] there using “face” filter and did get a lot of faces (along with a few non-facial images).
- BetaFace offers several ‘processing options’ including age, skin color, etc.
- Google images also employ face search (with imgtype=face added to the string) that seems to return most relevant results (well, no wonder). Google images has also recently introduced “photo content” image search.
- Face Search is a dedicated face search engine powered by Google search.
Image Search Based on Color
- Picitup allows to set color preferences (choose among 18 colors to set the search dominating palette);
- PicSearch and Snap.com recognize between colorful and black-and-white images.
- Etsy searches only inside its own product database but its color-based search engine is both fun and pleasure to play with.
Image Search Based on Similarity
Recently launched Tineye.com (registration required) searches for similar images online. While the database size still leaves much to be desired, this type of image “reversed” search is a great move. Like Etsy color search, that’s one of those search types requiring no keywords at all.
Images continue to be a vital part of creating quality content on the internet. These search functions allow us to find images that meet our needs – whether that is for shopping or content discovery. It will be interesting to see how these tools evolved as search options expand.
Featured Image: artjazz via Depositphotos
In Post Image: Deposit Photos