This article was 1st published on our sister Site, Digital World Native.
Streaming video Online takes up lots of expensive computing & network resources, & a user’s experience can be really poor because of a video being pixelated or due to buffering. Now MIT has developed ‘Pensieve’, an Artificial Intelligence (AI) system run on machine learning that picks the optimal algorithms based on network conditions. A higher quality streaming experience is now possible with less rebuffering than existing systems.
The algorithms employed by popular streaming services like YouTube or Netflix break videos up into small chunks which get loaded up as you watch. When your Net connection is slow, your resolution may get reduced to make sure you still get an uninterrupted viewing experience, which is why you’ll experience pixelation. When you skip ahead further into the video you start a rebuffering process & you may suffer from pixelation again.
These adaptive bitrate (ABR) algorithms save bandwidth & give users a consistent viewing experience, while saving bandwidth. The trade-off is that perceived video quality can often be low, resulting in lost revenue when viewers prematurely abandon clips, so researchers have been looking for better, cheaper alternatives.
For the rest of the article, click here.
