With the rise From live streaming, the game has evolved from a toy-like consumer product to a full-fledged legitimate platform and medium for entertainment and competition.
Twitch’s viewer base alone has grown from an average of 250,000 concurrent viewers to over 3 million since its acquisition by Amazon in 2014. Competitors like Facebook Gaming and YouTube Live are following similar trajectories.
The explosion in viewership has fueled an ecosystem of support products as today’s professional streamers push technology to its limits to increase the production value of their content and automate the repetitive aspects of the video production cycle. .
The biggest streamers hire teams of video editors and social media managers, but growing, part-time streamers are struggling to do it on their own or find the money to outsource it.
Online streaming play is a chore, with full-time creators performing eight or even 12 hours a day. In an effort to capture viewers’ attention, 24-hour marathon streams are not uncommon either.
However, those hours in front of the camera and keyboard are only half of the streaming stream. Maintaining a constant presence on social media and YouTube is fueling the growth of the streaming channel and attracting more viewers to watch a live broadcast, where they can purchase monthly subscriptions, donate, and watch advertisements.
Distilling the most impactful five to 10 minutes of content over eight or more hours of raw video becomes a sizable time commitment. At the top of the food chain, the biggest streamers may hire teams of video editors and social media managers to tackle this part of the job, but growing, part-time streamers struggle to find the way. time to do it themselves or find the money to outsource it. There are not enough minutes in the day to carefully examine all the images besides the other priorities of life and work.
Computer vision analysis of the game user interface
One emerging solution is to use automated tools to identify key moments in a longer broadcast. Several startups are competing to dominate this emerging niche. The differences in their approaches to solving this problem are what differentiates competing solutions from one another. Many of these approaches follow a classic computer dichotomy between hardware and software.
Athenascope was one of the first companies to implement this concept on a large scale. Backed by $ 2.5 million in venture capital funding and an impressive team of Silicon Valley Big Tech alumni, Athenascope developed a computer vision system to identify key clips in longer recordings.
In principle, it’s not that different from how self-driving cars work, but instead of using cameras to read nearby road signs and traffic lights, the tool captures the player’s screen and recognizes them. indicators in the game user interface that communicate important in-game events: wins and deaths, goals and saves, wins and losses.
These are the same visual cues that traditionally inform the in-game player what is happening in the game. In modern game user interfaces, this information is high contrast, clear and unobtrusive, and usually located in fixed and predictable locations. on the screen at all times. This predictability and clarity lend themselves extremely well to computer vision techniques such as optical character recognition (OCR) – reading text from an image.
The stakes here are also lower than in self-driving cars, as a false positive from this system produces nothing more than a less exciting than average music video – not a car crash.