<img src="https://d5nxst8fruw4z.cloudfront.net/atrk.gif?account=D6K+n1QolK10/9" style="display:none" height="1" width="1" alt="">

Your Auto-tagging Recipe for Success

by Jake Athey, December 15, 2017

Auto-tagging

Artificial intelligence, once the invention of science fiction, is quickly changing the status quo for entire industries and businesses, including the world of digital asset management. DAM users of today can now use image recognition technology to help relieve the burden of manually tagging their images. There are many auto-tagging integration options available to DAM vendors, but at Widen, we’ve approached things with our very own special spin.

Here’s what we cooked up for you!

1. Not one, but four options

For over five years, we’ve been experimenting with leading image recognition services. In fact, this summer, we asked 49 of our customers to test drive various technologies and share their feedback. We found that their sentiment matched much of our own, in that no one technology provider reigned supreme.

As a result, we integrated our DAM system, the Widen Collective®, with not one, but four image recognition services – Amazon Rekognition, Clarifai, Cloud Sight, and Imagga. We then give you, and other fellow customers, the flexibility to choose the technology or combination of technologies that best meets your unique business needs and use cases.

2. Advanced recognition capabilities

Because we integrated a variety of image recognition services, you have access to more capabilities than would be possible with just one. Currently, the services, as a whole, detect:

  • General keywords: Automatically add basic descriptive metadata for concepts and objects like stock photography or general lifestyle images.
  • Sentiment: Analyze facial attributes to auto-tag images based on emotions like happiness, surprise, and confusion.
  • Demographics: Categorize images of people based on age, ethnicity, and gender.
  • Not Safe For Work (NSFW) content: Auto-delete, unrelease, or restrict permissions for questionable or inappropriate images.
  • Colors: Auto-tag and categorize uploads based on image colors.
  • Industry specifics: Receive industry-relevant tags by tuning your service to an applicable business category, such as apparel or food.
  • Celebrities: Identify celebrities by name based on facial detection.
  • Product specifics: Identify and tag general and brand-specific products.

Each solution excels at different areas of image recognition, so depending on your priorities and the types of metadata you require, you can handpick the service or services that will help you reach your goals. Here’s how each of the four options stack up:

Auto-tagging

3. Managed integrations with the customization and flexibility you need

If you decide to leverage our auto-tagging solution, we’ll manage the integration from start to finish. In other words, we do all the heavy lifting while you tell us what to lift! Of course, we’ll guide you through the process, starting with a discovery call, and from there handle the setup, configuration, and maintenance, all while customizing the integration to your specific needs.

Here are a few examples of customizable configurations:

  • Security: Require final admin approvals or restrict which users have access to auto-tagging altogether.
  • Categorization: Tag or categorize assets into different groups or specific collections, based on specific keywords.
  • Inclusivity: Roll out image recognition services universally or limit to particular asset groups, such as lifestyle images only.
  • Quantity: Specify the maximum number of tags generated per image based on quantity caps and confidence scores.
  • Lists: Use whitelists and blacklists to account for definitive inclusions and exclusions.
  • Dependencies: Trigger if-then scenarios to maintain tagging consistency.
  • Notifications: Post errors or messages automatically to collaboration tools like Slack.

Now what?

We believe that humans are still needed (admins, we’re talking to you!). AI tools, like image recognition, require a balance between automation and human touch. While the goal is that technology like this frees up time and space for other value-driven projects, YOU are still needed to inform and guide both the technology and its users.

So, let’s start with a human-to-human conversation. Contact us to talk with your Widen advisor or customer experience manager to cook up an integration that fits your unique needs or to experiment with these four image recognition technologies.

Topics: DAM, Integrations, Image recognition

comments powered by Disqus

The Widen Blog

Where marketing and creative teams find actionable advice, practical resources, and success stories to flourish in a world connected by content.

Like what you're reading?

We can send you our best articles.

Workflow Workbook

This creative workflow workbook is intended to help you start the dialogue with your teams to figure out what’s going on with your current creative workflows and how your teams can start producing better work together.

Get workbook

 

Research Report

2017 Widen Connectivity Report

See what 221 marketing, creative, and IT professionals have to say about how they want to connect today, what impact connectivity has on their work, and where they see connectivity going in the future.

Get the free report

Recent

Are you ready for a DAM solution of your own? Get Started