Manual tagging of metadata for digital assets like images and video is time-consuming, tedious and in most cases incomplete.
With Cordiaa in place, your digital assets are automatically tagged at scale using Computer Vision. This means images and video segments can be found very easily, increasing cycle times and enhancing productivity and efficiency.
Cordiaa detects any unsafe content uploaded by the users. Any inappropriate content that is flagged cannot be distributed within the organization thereby keeping your workplace safe and compliant at all times.
When you upload real-world images into Cordiaa, it automatically recognizes text in those images and tags them. This extends to videos as well where text on screen like ads, news, sports scores and captions are recognized and tagged.
As search indexes are auto-created based on the text present, you can do a visual search for images or videos containing a specific keyword text present in the image or the video.
Similarly, vehicles can be identified based on license plate numbers, products can be identified based on text-based logos and other information present on the images of the products and so forth.
Text in most Latin scripts and numbers embedded in a large variety of layouts, fonts and styles, and overlaid on background objects at various orientations as banners and posters can be identified, tagged and indexed for easy search and discovery.
For any uploaded image or video, Cordiaa recognizes the presence of people, objects and scenes in those resources and tags them accordingly. And for video, specific activities happening in the frame, such as “delivering a package” or “playing soccer” can be identified and tagged.
For example, in the street image shown here, the presence of persons, cars, buildings and other objects are recognized and the image is labeled accordingly.
Cordiaa has implemented a hierarchical taxonomy of ancestor labels to categorize labels. So, if a person walking across a road is detected as a Pedestrian, the parent label of Pedestrian which is Person is returned as a label as well and tagged accordingly.
Taken this way, if you ask the Bot on Cordiaa, “Show me all images which have vehicles in them”, the Bot might deliver all images with cars, motor bikes and any other vehicle in an image.
When you upload an image that contains a face or a number of faces, Cordiaa detects each of the faces in the image, analyzes the facial attributes of each, and then tags the facial attributes that are detected in the image.
This includes gender (male, female), approximation of the age of the person (23-35 years old), emotion (appearing happy or sad), other facial attributes like whether smiling, whether wearing glasses, whether wearing sunglasses, whether having a mustache and so forth.
Cordiaa supports the creation and management of one or more Face Collections. For instance, you can create a Face Collection for your employees. Or, you may create a Face Collection for your customers. Each of these will map to a Face Collections data store which can be referenced for querying images and videos for face comparisons.
A Face Collection of employees can be referenced for providing access to a facility where the image from the live camera at the entrance is compared with the reference image in the Employee Face Collections and access given or denied to the facility.
Or you could simply ask the Cordiaa Bot: “Show me all the images that have Jane Doe”, and Cordiaa will list all images which have Jane Doe.
You can also create and configure Ingestion Channels on Cordiaa where you can harness all the valuable content captured on mobile phones and other devices through a familiar chat interface.
For example, if you are a consumer products company and have people on the field collecting market data, you can now elevate their work to the next level by having them shoot more photos and videos of your products as well as your competitors’ products in the stores and have them all uploaded directly from their phones into a Cordiaa channel specifically created for that purpose. These assets are then automatically tagged for easy search, discovery and analysis.
You can also direct live images of shoppers in your stores directly into a Cordiaa channel from where the images can be analyzed and retrieved based on emotion and demographic detail.