Image & Video AI Recognition

With Ci's image & video Recognition feature, users can automatically categorize, tag, and search for images and videos based on their contents, enabling efficient organization and retrieval of media assets. While these media files are processing, AI will identify objects, scenes, faces, text, and other elements within images and videos with remarkable accuracy. This functionality enhances workflow efficiency, simplifies content management tasks, and unlocks new possibilities for content discovery. In this article, we will outline how to use this feature so you can become an AI Recognition professional. 

Request AI Recognition from Customer Success

As of today, AI Recognition is only available for Company Network accounts. For any user of a Company Network that is interested in enabling this feature, please reach out to the Ci Customer Success team by submitting a request here Screenshot 2024-01-22 at 3.43.47 PM.png

Please note that this feature is enabled at the folder level. We recommend using distinct folders if you plan to use both Image & Video Recognition.

Image Recognition

Once we've enabled image Recognition for a folder, navigate to that folder in the Workspace and upload an image file directly into it:

Screenshot 2024-04-04 at 4.54.49 PM.png

Figure 1: Uploading an Image File

After the file finishes processing, AI Recognition will begin scanning the file to identify elements. This process can take some time. Once completed, select the Overview panel and click the Metadata tab to view the objects that were tagged to the image: 

Screenshot 2024-04-04 at 4.56.16 PM.png

Figure 2: Auto Generated Object Tags (image)

Video Recognition

Video Recognition works the exact same way. Once we have enabled video Recognition for a folder, navigate to that folder in the Workspace and upload a video file directly into it. After processing & AI scanning is completed, you will be able to see the objects that were tagged to the video: 

Screenshot 2024-04-04 at 5.08.56 PM.png

Figure 3: Auto Generated Object Tags (video)

Metadata Search

Now that these tags are applied to your files, you can utilize metadata search to locate files with ease. Start by selecting the Search icon near the upper-left corner of the UI: Screenshot 2024-04-04 at 5.13.54 PM.png. Within the Search modal, select the Search Options tab (see figure 4). At this point, the metadata name field will need to contain one of the following:

auto-gen::Objects (if searching for objects)

auto-gen::Objects/Activities (if searching for activities)

auto-gen::Celebrities (if searching for celebrities)

In the figure below, we are using auto-gen::Objects in the metadata name field. We will also use "Mountain" as one of the values to search for as well. Next, hit Search:

Screenshot 2024-04-11 at 1.36.54 PM.png

Figure 4: Metadata based Search

On the following page, you will discover all the files that match the search criteria. 

Screenshot 2024-04-11 at 1.47.25 PM.png

Figure 5: Results page

Like other queries made in Ci, the search results page provides a number of related facets to help navigate through the results. You can filter by:

— Network

— Space

— File type (video, audio, document, folder, image, etc.)

— Archive status

— Aspect Ration

— Audio Sample Rate

— Video Codec

— Frame rate

— Pixel Dimensions

— Runtime

To narrow down search results, the user should simply click on a facet fag. 


For any additional questions, please reach out to the Ci Customer Success Team by submitting a request here Screenshot 2024-01-22 at 3.43.47 PM.png in the Help Center.

Have more questions? Submit a request