Metadata

Adding quality metadata to your digital artifacts is among the most challenging aspects of a large-scale digitization project. Digitizing a collection is only half the battle. Digital artifacts are often worthless–both for historical and commercial purposes–without clear metadata describing what each artifact depicts, when it was created and who created it. Descriptive titles, locations, keywords and other metadata further enhance digital artifacts, making them more accessible to scholars and researchers and easier for buyers to find on commercial marketplaces.

What We Do

Gado Images’ innovative Cognitive Metadata Platform™ (CMP) is a proprietary cloud-based software system which uses groundbreaking technologies like facial recognition, optical character recognition, and neural networks to automatically caption and keyword images. Integrated with services including IBM’s Watson, Google’s Vision API, and Amazon’s Mechanical Turk, our Cognitive Metadata Platform™ allows us to process large volumes of material quickly and increase the searchability and commercial value of your digital assets.

Auto Tagging

Gado Images’ Cognitive Metadata Platform (CMP)™ reads the visual and textual content of your digital images, automatically tagging significant people, locations, events and objects. Standardized with controlled vocabularies including DBpedia and Getty Images, Gado Images’ CMP automatically exports structured semantic metadata about your images, greatly enhancing the searchability and commercial value of your content.

The Cognitive Metadata Platform™ can even automatically write a well-formed, sentence-length, human-readable caption for each of your digital artifacts based on the people, locations, events and objects it finds. Other capabilities include automatically estimating dates for historical images and automatically detecting photographic compositions.

The CMP can further enhance its automatic capabilities by ingesting existing metadata about your images, or can automatically commission new metadata through Gado’s distributed Amazon Mechanical Turk team or our US-based team of professional image researchers.

Facial Recognition

Gado Images’ Cognitive Metadata Platform (CMP)™ uses advanced facial recognition technologies to automatically find and tag significant individuals in your images. The CMP draws on a pre-trained database of over 60,000 well-known individuals, and can learn new individuals in your organization’s collection.

Gado Images’ CMP also uses facial recognition to automatically determine the number of people in your images, estimate their age and gender, determine and tag their emotions and facial expressions, and determine the photographic composition of your images.

Amazon Mechanical Turk

Gado Images’ Cognitive Metadata Platform™ integrates with Amazon Mechanical Turk (MTurk), a crowdsourcing platform from Amazon.com that uses a distributed workforce to perform tasks which computers are unable to do.

CMP’s MTurk integration allows the platform to automatically add human-written keywords, titles and other metadata to your images. All MTurk keywords and other metadata are automatically standardized to CMP’s controlled vocabulary, allowing the platform to add high-quality, human-written metadata to your images in a matter of minutes, and at a very competitive cost.

IBM Watson

IBM’s Watson system is best known for beating several human champions on the television show Jeopardy!. Within Gado Images’ Cognitive Metadata Platform™ however, Watson provides more practical capabilities, including advanced natural language processing, sentiment analysis, relationship extraction and computer vision.

Just as Watson learned how to play (and win) Jeopardy!, CMP’s Watson integration allows the platform to learn about the contents of your collections, lend structure to your existing textual metadata, and identify relationships between the people, places, events and objects in your images.

Optical Character Recognition

Many digital artifacts include written content, in the form of attached documents, newspaper clippings, or even handwritten notes from an original photographer or artist. This written content provides valuable information about the artifacts, but transcribing it is often time-consuming and expensive.

Gado Images’ Cognitive Metadata Platform (CMP)™ uses powerful Optical Character Recognition (OCR) technologies to automatically read the written content of your images and transform it into searchable text. This text can then be used with CMP’s auto-tagging capabilities to automatically tag people, places, events and objects in your images.

From transcribing text off street signs in your photographs to quickly processing pages of newspaper clippings, CMP’s OCR capabilities automatically add an extra layer of searchable, machine-readable metadata to your collections.

Google Vision

Google is a pioneer in machine vision. Through its industry-leading search engine, the company indexes billions of images every year, and uses those images to train sophisticated algorithms which can recognize objects in photographs, find text, and perform other functions.

Through Google Vision, Gado Images’ Cognitive Metadata Platform (CMP)™ taps into the power of Google’s algorithms, allowing the platform to automatically find important objects, landmarks and other entities in your images.

Image Formatting

Getting images into the right format for commercial distribution or online distribution can be a challenge. Gado Images’ Cognitive Metadata Platform (CMP)™ accepts nearly any type of image file, and automatically formats your images according to industry standards. Functions include auto cropping, auto rotation, intelligent resizing, automatic grayscale/color detection, watermarking, and IPTC/EXIF extraction.

In addition, CMP provides valuable insights about the photographic content of your images, such as dominant colors and visual complexity, which makes your images more searchable and valuable on commercial marketplaces.

Getty Integration

Getty Images’ Enterprise Submission Platform (ESP) allows large-scale contributors to automatically send content from their own systems directly to Getty Images, along with structured metadata. ESP also allows contributors to track their content through each step of Getty Images’ approval and publication process, and to see insights on how their content is used.

Gado Images’ Cognitive Metadata Platform (CMP)™ is fully integrated with Getty Images’ ESP. When Getty Images debuted ESP, Gado Images was chosen as one of only two contributors given early access to the system, and the Gado Images team works closely with Getty Images to test new features and provide feedback on ESP changes. The result is faster approval times for your content, fewer metadata errors, and better search performance on the Getty Images site.