Vionlabs Press Release 2021 Feb 2

Vionlabs is significantly expanding its Video Fingerprint API portfolio tackling the discovery problem experienced by subscribers today


By Vionlabs


PwC’s REPORT shows that 68% of the households maintained their cable TV subscription while keeping multiple streaming subscriptions. A market research done by Vionlabs and Dynata in 2021 among 18-64 years old in the U.S. found out that 40% of people have more than 3 subscriptions at the same time.

The winners of the streaming wars will be the services that can create an unique and differentiated user experience, Vionlabs expanded the portfolio to help streaming services to achieve that goal and stand out from the competition.

The new expanded portfolio focuses on solving two problems:

  • Minimize frustration -> minimizing the frustration felt by users when trying to find something to watch. 30% of the users say that they spend more than 21 minutes searching for content each time.The portfolio that solves this challenge includes everything in discovery such as recommendations, search, editorial and visual aspects such as previews, posters/thumbnails and content descriptors

  • Maximize Entertainment -> Maximizing the entertainment for users while consuming their favorite content. 55% of users say that “easy-to-use” is important for them to stay with the services. The portfolio that enhances a good watching experience includes binge markers (intros, recaps and end credits auto-detection), optimal ad insertion points and more emotionally matching ads

Vionlabs generates the data through analyzing audio and video files, powered by AI, independent from any 3rd party data. The Vionlabs services can be easily accessed through APIs, and parts of the expanded portfolio are already deployed in the market by known names such as Tele2 Play, Canela Media and more.

Fingerprint+ - Enabling relevant and diverse content discovery

With Vionlabs Fingerprint+ you will get:

  • High-dimensional Content similarity embeddings. Proven to increase recommendation accuracy by up to 83% and catalogue coverage by up to 56%

  • Moods (Ex. Dark Crime, Emotional & Inspiring, Lighthearted & Funny)

  • Genres (Ex. Action,Comedy, Drama, Biography)

  • Video Story Descriptors (Ex. Thrilling, Survival, Underdog)


All of this is 100% extracted directly from the video file with Vionlabs proprietary AI-networks, trained to handle movies, series, documentaries, reality-tv and much more. With fingerprint+ you no longer have a coverage problem nor a quality problem for your content catalogue: 100% coverage of the highest quality data, enabling your entire library to be discoverable for your users with better search, better recommendations, better curated lists, mood lists and much more.



Visual Discovery - Accurate representation of the content for increased engagement

64% of users in our latest research revealed that they had passed on a title due to the poster being offsetting, but later found out that this was something they enjoyed. To solve this problem, Vionlabs is now launching its Visual Discovery portfolio, using Video deep learning to automatically generate:

  • Automated previews with mood tags

  • Automated thumbnails with mood tags (coming soon)

  • Automated Posters (coming soon)

Binge Features - Maximize the entertainment viewing experience for users

Almost 50% of the users nowadays binge watch series very often (watching 2-3 espides in one sitting) and 93% of the users say that they want their services to have binge watch features. The Vionlabs binge portfolio includes:

  • Automatic detection of intro, recap and end-credits

  • Automatic detection of optimal ad placement (beta)

  • Emotional time series analysis for ad-insertion (coming soon)



About Vionlabs

Vionlabs was founded with an end goal to minimize frustration and maximize entertainment for all viewers, by creating high quality data for video content. The data helps media businesses to make better decisions throughout video acquisition, production and distribution processes. It addresses this through a series of highly innovative AI-networks that use multi-layered deep scene-by-scene content analysis to create the industry’s most advanced analysis of video content, only by analyzing audio and video of the content.

39 views