Google has been granted a patent for "Automatic large scale video object recognition". The technology described would enable Google to recognize objects (not just faces) in videos.
The abstract explains:
An object recognition system performs a number of rounds of dimensionality reduction and consistency learning on visual content items such as videos and still images, resulting in a set of feature vectors that accurately predict the presence of a visual object represented by a given object name within an visual content item. The feature vectors are stored in association with the object name which they represent and with an indication of the number of rounds of dimensionality reduction and consistency learning that produced them. The feature vectors and the indication can be used for various purposes, such as quickly determining a visual content item containing a visual representation of a given object name.
Obviously, YouTube is the first thing that comes to mind. Jon Fingas at Engadget suggests that Google would use the technology to fill in YouTube keywords without user intervention.
Some blogs are talking about how Google might be able to use the technology for Google Glass, identifying objects as we see them. I'm envisioning some Terminator-esque identification system (but honestly, probably much better).
Assume that people using Google Glass will be recording video of the imagery they see. We've already seen Google co-founder Sergey Brin touting the instant uploading of the device, with it uploading pics from every ten seconds.
This could also lead to greatly improved visual search. Think Google Goggles, and the advancements Google has already made. With a recent update, Google will now add search results via notification when it simply recognizes something you're taking a picture of with your Android camera (even without actively using the Google Goggles app).
The more Google can identify objects visually in this way, the better it will be able to associate search results with them. The fact that Google is already reducing its dependence on keywords for search would only add to the potential.
It wouldn't do anybody any good to speculate too much on what Google might have in store for this technology, but it certainly gets the imagination going. The image of the boots above certainly makes me consider the product search implications.
You can read the actual patent here.
Images: Terminator 2: Judgment Day