Costanza Ai

The end-to-end cloud Si Media platform.

Migrate your system to the cloud to minimize your initial investments, to maximize the scalability and the flexibility of your solution.

What is COSTANZA?

What is COSTANZA?

Costanza is a platform composed by various applications, which together generate an end-to-end workflow for the management of contents in the cloud, from the acquisition of these contents, to the processing of them through our artificial intelligence platform, COSTANZA AI, up to the multi-channel delivery of the contents themselves, with COSTANZA FAST CHANNEL. 

yes costanza AI

Costanza Fast Channel is our cloud product capable of distributing video contents and stream signals to multiple destinations, linear and non-linear: to CDN, to OTT platforms, towards pay per view operators, with the possibility of integrating AD server to customize and target advertising content addressed to users.

  • Linear and nonlinear distribution
  • Integration of AD server

The Broadcast industry is highly automated, process-oriented and time-sensitive (especially for news and live productions). With Costanza AI, SI Media own artificial intelligence, you can automate all your media processes, from speech to text, face recognition, object recognition, to the most advanced AI features, such as Summarization, with automatic generation of highlights and synopsis.

Why adopting COSTANZA Fast Channel?

Diversification of
revenues:

Extend the offer of ad-supported video contents, by using dynamic\custom advertising and distributing the linear channels to connected devices (Samsung Smart TV, Chromecast, Roku), generating new revenues for the broadcasters.

AVOD (advertising supported video on demand):

Uploading and sharing of contents on demand, on OTT platforms, extracting further value from your own media library, either with subscribers or as a free access portal where to sell contents with a “premium” fee.

Reduced
implementation times:

Create a new channel, even for a single event, in few clicks, creating additional value from your existing infrastructure, without significant upfront investments, making your system even more flexible and agile.

Improved user
experience:

With this new fruition, user has a customize experience, where he sees only contents of interest (not only ADV, but also local sports or cultural events), increasing the engagement and the loyalty of the viewers to your television channels.

Highlights

Highlights

Catalogue and metadata generation

Automatic search and tag of content libraries, drastically reducing video process time.

On prem AI

Allowing televisions and public entities to keep their contents on prem while processing them.

Improve your live production

Automatically generate highlights from sports events and synopsis with AI summarization.

Live censorship

Identify explicit contents such as nudes, alcohol, tobacco through AI.

What’s in our COSTANZA AI?

What’s in our COSTANZA AI?

Face
Recognition

Detect all faces inside a video track, including not recognized ones. Able to recognize faces recurrent in more videos, with a dedicated interface to edit/assign names to the detected faces. Optimized search on asset archive and inside the asset for faster detection of faces.

Object
Recognition

Detect objects inside a video track based on the desired objects to search. Model can be customized with the desired prompts, with an optimized search on asset archive and inside the asset for faster detection of objects

Content
Moderation

Detection of unsafe contents in a video track based on category. Optimized search on asset archive and inside the asset for faster detection of unsafe assets. Supported categories including nudities and explicit contents (tobacco, alcohol).

Speech to text
& translation

Transcription and translation of the speech in a wide range of languages. Automatic subtitle generation in .SRT format and optimized search on asset archive and inside the asset for faster detection of spoken words.

Semantic
Search

Advanced search on asset archive of scenes automatically extracted from the video, based on the text prompts given as input, understanding the contextual meaning and intent behind a user's search query.

Summarization

Automatic synopsis generation based on the Speech to text result. Automation highlights generation from a sports event video.

Posters

Automated selection of a significant and pleasant frame from a video and generation of posters based on a predefined template, choosing the best cropping area, based on each desired aspect ratio. Dedicated interface for defining different templates and poster formats.

Sentiment

Analysis of emotions in a video in order to detect positive and negative moments. Optimized search on asset archive and inside the asset for faster detection of positive and negative assets. Supported categories: Neutral, Sad, Angry, Happy, Disgust, Fear, Surprise.

Book a DEMO

Fill in the form below for more information and we will be happy to help you and advise you on the best solutions.