Google and Kaggle have announced today a new machine learning challenge where they ask developers to find the best way to automatically tag videos.
The challenge comes with a $30,000 grand prize, (and $25,000, $20,000, $15,000 and $10,000 for the next four teams), asks developers to classify and tag videos from YouTube. The data set they are working with includes a total of 7 million YouTube videos that adds up to over 450,000 hours of video. YouTube-8m includes labels that developers can use as their training set. The challenge then involves them tagging 700,000 previously unseen videos.
The company is launching this new challenge on the same day of Tensor flow’s 1.0 release which is by no means a coincidence. Google doesn’t restrict developers to using its own machine learning frameworks though, as they are free to use whatever they wish. Given that the full frame-level data is over 1.7 terabytes chances are most developers will want to use Google’s own services to train their models (and they can get a few extra free credits to use the Cloud Platform, too, once they run out of their free allotment).
This is definitely an interesting push from the company as they push for machine learning and video classification. These classification tools will likely be used in both YouTube and Google’s search engine to improve results for their users.