Aims to develop a photorealistic 3D human avatar for live sign language interpretation of television programmes

Content4All enables the low-cost personalisation of content for the deaf community without interrupting standard programming by developing technologies and algorithms to achieve automatic sign language translation.

The Content4All project aims to make television content more accessible to deaf viewers by developing in studio a 3D human avatar that can sign to be used during TV shows.

  • Extraction of sign language interpretation parameters, real-time development of 3D models and animation.
  • Evaluation of trial performance and development of sustainable business models
  • Development of sign language interpretation signal
  • Development of a marketable demo

CONTENT4ALL was cited in Gartner’s "Market Insight: Three Key Areas in Which AI Will Disrupt the Video Entertainment Ecosystem” report as one of the AI based projects that is most likely to radically revolutionise the creative video industry. It is in fact a highly innovative project.

Innovation features

  • Rendering / photorealistic animation of human avatars:

    a new framework for low-cost, low-complexity acquisition and animation for automatic translation into sign language.
  • Animation based on photorealistic images:

    a new engine that only draws from a limited database of images and video of core expressions
  • Automatic connection and fusion of sequences of movement and core expressions

    to form an entirely new video sequence
  • TEffective translation of sign language:

    this new approach combines a static and a rule-based approach to overcome the lacunae of both systems individually
  • HbbTV applications and synchronisation of content for the development of dynamic cloud-based content: 

    innovation to guarantee translation signal reception latency in real-time does not impact user experience.
  • Sets a standard for user interfaces designed for users with disabilities:

    defines usability and UX for deaf users. Identifies the essential features of content, of transmission and of the user interface.