These auto-mixing features will be included
in video switchers in the future to allow for a
completely AI-switched production. It will eventually replace the role of a technical director for
Computer-vision-based video switchers can
work independently on embedded systems or
devices on-premise using existing hardware.
Cameras can even leverage a networked cloud
server if needed.
Creating Automated Actions
and Triggers for Real-Time Graphics,
Animations, or CG Characters
A neural network can identify target objects
or people with facial recognition, which can
trigger production events such as generating a
lower-third for a presenter at a conference. Facial recognition could also generate graphical
statistics on a particular player on the field, or
even allow control of a CG character to be inserted into a stream.
Cognitive technology will be prevalent in everything—sports, eSports, corporate communications, education, and live events. This will integrate data-driven assets and visualizations that
change according to specific actions, times, locations, or dynamic data in relation to the stream.
Natural Language Processing (NLP) allows
for automated live transcription, translation, interpretation, captioning, and audio description
for use in meetings, lectures, or events. This
would be useful for multinational corporations
that need live captioning for town halls, product
launches, or general communications in multiple languages for a worldwide audience.
Video Analytics and Metadata
Extraction for Data Management
As companies get much more involved with
streaming, the sheer volume of data generated
from video is increasing exponentially. The infor-
mation derived from this data can be leveraged
beyond what humans can extract manually.
AI will interpret streaming content and extract metadata by generating descriptive tags,
categories, and summaries automatically. This
will allow for more intelligent analytics, content insights, and better content management,
paving the way for efficient methods of monetizing video through targeted ads.
Monitoring Social Media
Sentiment and Auto Sharing
Social media monitoring allows brands to
gauge online conversations and sentiment analysis, tracking audience reaction in real time.
This allows for immediate customization or adjustment of the content to suit the audience’s
stated preferences. Natural language algorithms
will pull data from the stream and capture major
topics and keywords, and then compile screen-shots, video scripts, and highlight clips that can
be used for marketing purposes or automatically uploaded to social media.
Artificial intelligence will be a powerful tool
for companies in the streaming industry to leverage once they are able to unlock its full potential.
We have just begun to scratch the surface of the
full power of AI for intelligent streaming. The examples above are just a small sample of how AI
can enhance live streaming by making live content more engaging and efficient, as well as allowing cost savings from production to delivery.
AI will propel content owners, media producers,
and advertisers into a new creative mindset of
making smart and compelling content.
Mark Alamares ( email@example.com) is a tech entrepreneur and
media strategist for Fortune 500 companies. He is the CEO of
Mobeon, an advanced media studio and consultancy for enterprise,
specializing in immersive content production and distribution.
Comments? Email us at firstname.lastname@example.org, or check
the masthead for other ways to contact us.