Please wait while we enable your Account


Contacting Amazon Web Services
Deploying Cloud Servers, Storage, Transcoding & Database Servers
Deploying Global CDN
Deploying Firewall & Enabling Security Measures
Deploying the CMS & Admin Module
Deploying Website, Mobile & TV Apps framework
Creating your FTP account
Finishing up all the modules
Preparing for launch

Video Streaming buffering might be prevented by Artificial Intelligence

Roshan Dwivedi Published on : 17 August 2017

According to Massachusetts Institute of Technology (MIT), Artificial Intelligence could be the answer to reducing headaches for viewers and video streaming services. Video buffering and pixelation continues to be problematic for those who rely a ton on video streaming services for catching up on the newest movies and the latest TV shows.

Buffering and pixelation can make people to switch over to other content which can result in poor viewership and advertising. The way it works is that Streaming services use ABR (Adaptive BitRate) algorithms to ascertain what the resolution of the video playback is at that time which varies according to changing network conditions. Videos are typically chopped into smaller chunks and transmitted in a sequential manner to the device that viewers are watching it on. Algorithms also work to create a buffer ahead which means that the streaming experience will be uninterrupted even for a short period of time.

Changing network conditions also impact the streaming causing sudden pixelation of the videos being watched which in turn leads to long pauses, especially if you skip the video too far ahead.

MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) has come up with “Pensieve” neural network, which is an artificial intelligence (AI) system. This system uses machine learning to pick up different algorithms to adjust to the varying network conditions.
This enables the easy switching between rate-based algorithms or buffer-based algorithms as per network conditions. This predictive technology lets streaming priorities to be set. So if a user is watching a video on YouTube and is going to walk into an area where there is poor connectivity, this intuitive system will be able to reduce the streaming resolution in an appropriate manner so as to buffer only for potentially falter free streaming. Netflix and Amazon Prime Video have consistently optimized after improvements in streaming algorithms to offer optimal viewing experience on slower or inconsistent mobile and home broadband networks. However, network condition variations still play havoc from time to time. So MIT’s “Pensieve” neural network could be the answer to these issues thus enabling streaming higher resolution 4K content easier for streaming platforms.

Written by: Roshan Dwivedi

Add your comment

Leave a Reply

Your email address will not be published.

Try Muvi One free for 14 days

No Credit Card Required
Your website will be at, you can change this later.

Upcoming Webinar
October 11

9:00AM PST

Creating a New-Age Customer Support System

Creating a new-age customer support system requires leveraging modern technologies and strategies to provide exceptional customer service which includes creating workflow laced tools for listing and managing customer enquiries, issues, internal requests, delivery backlog, release management, feedbacks, To Dos for teams, SLAs etc seamlessly by tailoring it as per your processes and policies. In this webinar you get to understand how to implement a strategic customer support system by understanding our innovative workflow management tool and also managing integrations with ease. Also it will help you be able to track, monitor the deliverables in an effective and efficient manner. 

In this webinar, you will learn :

  • Tattle’s Architecture Unveiled: Discover the backbone that powers your deliveries and processes.
  • Tool Creation Process: A detailed walkthrough how to create tools for customer , Internal teams, manage workflows and SLAs
  • Integrations: Effortlessly integrate with multiple data sources to collect and manage enquiries, issues, feedbacks  

Event Language: English
1 hour