Machine Learning Masterclass #6 Machine Learning and Video
In our Machine Learning Masterclass series so far, we have covered what Machine Learning actually is, the impact it might have on SEO and the sorts of tools and applications that are currently powered by Machine Learning. In this blog post, we want to delve in to see how Machine Learning is having an impact on the ever-popular medium of video, and what could change in the future.
As we have discussed in previous blogs, Google is leading the way when it comes to the development of artificial intelligence (this is part of the reason we think machine learning will have such an impact on search). Fittingly, just last week Google announced that it has released Cloud Video Intelligence API, in private BETA. This is one of the most groundbreaking pieces of technological development when it comes to the impact of Machine Learning on video. Image recognition has been a significant part of Google’s Tensor Flow project for a while, but at Google’s Next Cloud conference, chief scientist of AI Fei-Fei Li made a significant announcement. The ‘Video Intelligence API’ is able to identify the content of the video without manual tagging, something that wasn’t previously possible.
What will be interesting is how Google can bring this sort of technological advancement into practical use. Tech companies are already demonstrating the possibilities, for example Snapchat are using Machine Learning alongside the video advertising that they host on the app. In a nutshell, the API allows greater ad targeting, by allowing advertisers to set goals on their campaigns beyond just impressions. It means they will have a greater idea which users are most likely to swipe on which type of ad. The fact a multimedia app such as Snapchat is invested in AI gives an idea of the impact Machine Learning has had, and will continue to have in the technology sector.
Unsurprisingly, mega video hosting platform YouTube has been using Machine Learning for the past year to produce its personalised recommendations for its users. This is something many big companies have introduced in recent years, including Amazon and Spotify. The recommendations API detects watching patterns automatically, so should keep up with interests week on week. Google recently released a competition that challenged people to develop classification algorithms that accurately assign video-level labels using the YouTube tagging system and the Google Cloud Platform. If nothing else, this proves the constant challenge of integrating Machine Learning with video. (If it was easy, everyone would already be doing it!)
In short, the crossover between video and Machine Learning poses a great challenge. Google’s recent Cloud Video Intelligence API is definitely a big step forward, and if Google can bring this sort of technology into public use then it is bound to be a goldmine. If you are curious to know how we see the future of Machine Learning, then check out the rest of our blog series. Remember, Machine Learning can affect your SEO rankings, so talk to one of our technical SEO experts today if you have any questions.