Google Can Now Automatically Tag Objects Within Videos
A new patent will mean that everything within YouTube videos can be instantly 'labeled.'
It's no secret that Google has been working on patents for facial recognition for quite some time, but now the leading Internet search engine has successfully dived into the world of object recognition. According to the abstract of the United States Patent 8254699, ‘an object recognition system performs dimensionality reduction and consistency learning on visual content items such as videos, resulting in a set of feature vectors that accurately predict the presence of visual objects represented by a given object name.' Google will use a database of ‘feature vectors' to label and automatically identify objects within videos, most likely on YouTube. The Google depository would hold at least 50,000 object names for the technology to automatically tag almost everything. The possibilities are endless for this recently granted patent, especially for marketers looking to promote their products.