MerlinOne announced NOMAD, their new technology to search the visual content of objects, with no dependence on textual metadata. Need a photo of four people at a picnic? Just ask NOMAD to find it for you. Organizations may have their photos, graphics, illustrations and videos in a Digital Asset Management (DAM) system, but they still cannot find the right object they need while on deadline for a campaign. The problem is the scarcity of useful textual metadata which, so far, has been the only way DAM users can search their collections.
By applying a combination of Deep Learning AI technologies (all developed in-house by MerlinOne) to your search query, NOMAD understands the concept you are looking for and delivers precise results based solely on the visual content of an image, completely independent of the existence or accuracy of its metadata.
NOMAD understands language patterns as well as visual content. Just like it knows to group visually similar images, it also clusters phrases of similar meaning, reducing the dependency on a specific, controlled vocabulary to find what you’re looking for. NOMAD can be combined with other MerlinAI tools such as IMPACT and Visual Similarity to provide additional search capabilities.