Home / Software & Service News / Machine learning could put an end to your streaming woes

Machine learning could put an end to your streaming woes

Whether you’re watching a YouTube video, streaming media, or playing games, the telltale sync signal associated with buffering can unload some severe existential dread.

Nothing takes you out of the experience and entertainment faster than a buffer circle continuously animated, seemingly to infinity. If you’ve fallen victim to buffering in the middle of an important moment in your favorite show or game, you’re not alone. More than 500 million hours of video are streamed per day which means the audiences of streaming services likely encounter a metric ton of buffering.

Buffer haters alike will be rejoiced to learn MIT researchers may have come up with a solution, that will hopefully end slow load times once and for all. They have developed a unique artificial intelligence system that can optimize video streaming, making the process reliable and devoid of animated icons.

To understand how it works, first, you must understand what buffering is and why it happens.

The tech behind the annoyance

Internet traffic or data is sent in units called packets, also referred to as chunks.

When it comes to video, streaming or the loading of said data happens in segments.

Over time, fragments come in as sequential portions of a whole file, which are then stitched together — that’s why you can often start watching a video with little to no interruptions but run into them soon after. At least to start, the content is played back to you at the same time it’s being downloaded.

During the process, if your connection drops or lessens, those chunks may stop flowing in, resulting in reduced performance for the entire file or video.

The idea is to continuously download these pieces and stitch the footage together in the background, as you watch. When the process is interrupted for whatever reason, you encounter buffering and the dreaded spinning circle.

This is exacerbated depending on the strength of a wireless signal, internet connection or nearby traffic. Watching a video on a public network that’s crowded, for instance, could result in significant buffering and performance issues.

Services like YouTube, Vimeo, and even social media platforms, rely on algorithms called Adaptive Bitrate, or ABR for short. They initially measure the connection speeds, total bandwidth availability, and the resolution of the content to deliver a constant flow or stream of media.

Higher resolutions, obviously, require more resources, so sometimes you can offset buffering by lowering the quality or resolution of the content — that’s why Netflix sometimes seems blurrier or worse than usual. The service or app has toned down the resolution of the content to match a higher connection demand.

So, how can this be solved through AI?

MIT’s solution

The team at the MIT Computer Science and Artificial Intelligence Lab (CSAIL) rely on an automated intelligence system to swap between the appropriate algorithms. The neural network — digital, of course — can analyze and decide when a connection requires one particular algorithm over another.

The team trained the AI system through a reward and penalty based system.

Over the course of a month, they played streams of video content and left the system to do its thing. Failures resulted in a penalty, while successes were rewarded.

This eventually allowed the AI system to work out which algorithms work best for various scenarios, and when it’s time to swap between them. They give it a virtual education.

Even more promising is the fact that the system can be adjusted and tweaked, depending on what a service, connection or media type calls for.

Content providers — such as Netflix — could opt to always select quality over performance or vice versa. The system would take this into account and make the right choices, through automation and regular monitoring.

The professor of the crew, Mohammad Alizadeh, says it can be completely customized. This would allow users to personalize “their own streaming experience based on whether they want to prioritize rebuffering versus resolution.”

This is great for haters of buffering sure, but the implications for other areas of media streaming are incredible.

Imagine being able to stream a high-resolution and intense gaming experience over VR using this system. Without it, it’d be nearly impossible because the continually shifting resolutions and qualities would cause some severe nausea and take users out of the experience.

This is pretty cool news, indeed.

Of course, the part about no more buffering is also exciting. Finally, streamers can say goodbye to that dreaded spinning circle, once and for all.

Kayla Matthews is a technology writer interested in AI, chatbots, and tech news. She writes for VentureBeat, MakeUseOf, The Week, and TechnoBuffalo.

Click Here For Original Source Of The Article

About Ms. A. C. Kennedy

Ms. A. C. Kennedy
My name is Ms A C Kennedy and I am a Health practitioner and Consultant by day and a serial blogger by night. I luv family, life and learning new things. I especially luv learning how to improve my business. I also luv helping and sharing my information with others. Don't forget to ask me anything!

Check Also

Elon Musk says Tesla Semi trucks are BAMFs


Elon Musk says Tesla's new Semi trucks are BAMFs.

Yes, for those of you at home, that's the widely known acronym for "Bad Ass Mother F***er."

Musk playfully threw out the acronym as a "scientific term" Tesla's development team used as they designed the new big rigs. The name sounds ridiculous — but the specs, if they wind up being legit, are actually pretty impressive.

Musk claims the Semi can accelerate from 0 to 60 mph in just five seconds, well faster than traditional diesel trucks. That's the cab by itself or with an empty trailer loaded on. Read more...

More about Tesla, Tesla Motors, Tesla Semi, Tech, and Elon Musk

Leave a Reply

Your email address will not be published. Required fields are marked *