ADEIA
the toughest R & D challenges the media can throw at us .”
Much of what Adeia does is underpinned by AI ( artificial intelligence ) and related to media , such as natural language processing ( NLP ) and computer vision .
Doken adds : “ For any company who ' s active in these fields – whether it ’ s for products they have today or they ’ re thinking of launching new products and services in new markets - Adeia will be there to tackle
Smart , immersive and personal content delivery It ' s a widely accepted fact that consumers gravitate to video services that are personalised . To achieve such personalisation , according to Doken , vast amounts of multi-modal data needs to be processed in real time and automated metadata needs to be created . Both would help enable faster searches , as a basic example , within a very vast library .
“ If I wanted to search for movies – for example , videos that were shot in London during the last 10 years that have Daniel Craig in it , that both my wife and I may like – that ' s not an easy task . If I were to ask the same question of my smart speaker , that ’ s an even more difficult query to interpret since it ’ s such a long , cascaded question . Our innovations , in practice , provide the best answer for the user in a very short amount of time ,” said Doken .
When dealing with content , privacy and security are also concerns that the whole industry is looking at . The shift from a linear TV world into a connected TV world will have significant ramifications for advertising . “ It was easier in linear TV because your cable company or the TV service provider had a lot of data about you , but , in a connected TV world , you ' re interfacing with many applications , and they may not necessarily be aware of each other . That ' s why you see a lot of consortiums in the connected TV advertising world , each trying to come up with standardised solutions so it ’ s easier to share standardised data – particularly in a post-cookie world .”
Doken believes first-party data , rather than third-party data , will gather importance . Contextual recommendations in terms of
technologymagazine . com 35