Visual Atoms

  • Peter NEYENS profile
    Peter NEYENS
    13 September 2017 - updated 3 years ago
Voting is closed. 2260 users have voted.
Country: 
United Kingdom

Society and industry generates massive volumes of valuable visual data – billions of images and videos – but we lack sophisticated recognition tools to discover the information we need. Visual Atoms has succeeded in developing visual analysis and search technologies enabling ultra-fast recognition and retrieval of any landmarks, locations, buildings, products, books, logos or objects without any prior training. We instantly uncover relevant information using only visual queries, without words and without the need for costly manual annotations. We call it visual artificial intelligence and it is changing the way we can interact with big visual data. 

About the Innovator

Visual Atoms develops unique solutions for the visual analysis of vast volumes of media, helping businesses save time and money achieving previously impossible results. Finding relevant content instantly in billions of images or video frames is an everyday problem for content creators/owners, police, security services, and developers of asset management systems. We deliver the capability to analyse and search vast volumes of images and videos using only visual queries, without words and without the need for costly manual annotations. 

/futurium/en/file/visual-atomsjpg-0visual-atoms.jpg

Visual Atoms
 Visual Atoms on Twitter 

What is the innovation

We have developed visual analysis and indexing technologies enabling ultra-fast search and recognition of landmarks, locations, buildings, products, books, logos or objects without any prior training. Our technology is robust to partial occlusions, illumination, etc. We supply an end-to-end visual search solution, interfacing with professional relational databases such as PostgreSQL and a REST API supporting on-premises or cloud-based SaaS deployment. We have systems successfully operating at multiple customer locations (TRL8). Our product line includes the FIND:IMAGE, FIND:VIDEO and GLIDE solutions. These advances were supported via the EU-funded BRIDGET project.  

Out of the lab – Into the Market

Our support from the EU's FP7 framework programme was invaluable, linking us with key stakeholders. It helped to define precise technology and product requirements, provided financial resources and gave continuous feedback. It critically supported our participation in trade shows (such as the IBC Future Zone) and ISO/IEC MPEG standardisation activities. It also provided constant focus on industry needs and helping us to differentiate our products from the competition.