Advanced-ai-desktop-ui

Enrich your brand assets with unique AI-generated tags
Enrich your brand assets with unique AI-generated tags
Customize AI models to meet your specific business use cases
Customize AI models to meet your specific business use cases
Adopt a smart AI approach to accelerate your business
Adopt a smart AI approach to accelerate your business

Enhance your business today with Image and Video Intelligence

Object detection
Object detection
Color detection
Color detection
Text recognition
Text recognition
Landmark detection
Landmark detection
Face recognition
Face recognition
Emotion detection
Emotion detection
Video analysis
Video analysis
Multi-lingual transcription
Multi-lingual transcription

AI That Fits Your Use Case

Artificial intelligence is rarely a one-size-fits-all solution. Every organization has a unique goal they’re hoping to achieve with AI and auto-tagging. MediaValet provides a smarter approach to artificial intelligence, offering three tiers of AI capabilities to address the auto-tagging needs of any business.

pre-packaged-ai-tier-1pre-packaged-ai-tier-2pre-packaged-ai-tier-3
Pre-Packaged AI
Customized AI
Specialized AI

Take advantage of pre-trained, out-of-the-box AI models to tag your assets

Train AI models to label your assets with tags customized to meet your specific business needs 

Work with data scientists to build AI models to address complex use cases that can’t be met using Customized AI

Pre Packaged AI Tier 1


Pre-Packaged AI

  • Enrich your assets with AI-capabilities driven by Microsoft Computer Vision
  • This is MediaValet’s basic AI option available to all customers
  • Tag photos with common object, color and text recognition capabilities

Example: Without having manually added any keywords to their massive library of photos, a tourism company can quickly and easily search for images of a “beach”, at “sunset”, with a “couple”.

Phone-Example


Customized AI

  • Generate auto-tags that address your unique use cases with Microsoft Custom Vision
  • Easily train your own custom AI-models and connect them with your DAM
  • Tag photos with unique SKUs, product names, employees, and much more

Example: A telecommunications company sells a variety of brands and generations of mobile phones. They train an AI model to differentiate between the various brands and generations of handsets, so they can search using terms like “iPhone 8” and “Samsung Galaxy S8”.

Tier 3 Bacteria Example


Specialized AI

  • Build complex AI models to address advanced tagging use cases with Azure Machine Learning Services
  • Work with data scientists to solve unique use cases, like identifying bacteria strains
  • Specialized AI is for use cases that can’t be solved using Customized AI

Example: A pharmaceutical company is performing a scientific experiment that requires the identification and tagging of strains of bacteria across thousands of images. As the differences between the bacteria require advanced knowledge to identify, the company hires a data scientist and programmer to build a fully-customized AI model. Once completed, Azure Machine Learning utilizes the new model to tag the new strains in a matter of minutes.

Ready to Talk AI and DAM?

Using one or multiple AI-tiers, MediaValet can help you to put proven artificial intelligence and machine learning technologies to work today, instantly improving your asset discoverability and addressing your unique business cases.

Let’s chat to discover which tier is right for you and help you get started!

 


A Smart Approach to AI

While artificial intelligence will aid your team in improving asset discoverability – if done right, it’s meant to assist, not replace manual tagging. With this in mind, MediaValet takes a unique approach to ensure that AI never reduces the effectiveness and accuracy of your DAM.

Our approach consists of 5 best practices:

Best Practice - 1
AI Generated Metadata Separated

AI-generated “cognitive” metadata should be kept separate from manually added keywords

When kept separate, users can easily toggle AI-generated tags (i.e. cognitive metadata) on or off during a search, enabling them to find assets that were previously undiscoverable.

Best Practice-2
AI Tags Best Practice 2
AI Providers Tracked Azure

AI providers should be tracked the same as DAM users

Tracking the actions of your AI providers or models is just as important as tracking the actions of users in your DAM. Auto-tags should be attributed to the AI provider or model that added them, so it can be audited in the future.

Best Practice-3
AI Generated Metadata Separated Manual Tags

AI-generated tags should be filterable

Enabling your users to easily search for an asset based on AI-specific filters, such as confidence level, AI provider or model, or image analysis date, improves asset discoverability and relevance for advanced users.

Best Practice-4
AI Generated Tags Convertible

AI-generated tags should be convertible

When AI providers and models generate tags with a high degree of confidence, you should be able to easily convert those tags into keywords. This improves search results when AI services are toggled “off”.

Best Practice-5
Best Practice 5 Object Recognition
AI Generated Tags Improve Over Time

AI-generated tags should improve over time

Each action taken on AI tags should act as negative or positive feedback for the AI models your using – consistently improving accuracy.

Download Our Free Guide!

Read our free Guide to Enhancing Asset Discovery with Advanced AI
for a step-by-step implementation plan.

ai-brochure-mediavalet

  • This field is for validation purposes and should be left unchanged.