Image tagging
Talk to an expert
Talk to an expert
Need
Enable search engines to easily retrieve correct images from archives containing tens of millions of photos, overcoming the limits of manual tagging and poor metadata quality.
Approach
We developed a custom image analysis engine, based on neural networks, that automatically recognizes and weight visual content and keeps learning over time, so it stays aligned with new subjects, trends and contexts.
Result
A highly informative structured database that dramatically improves search relevance and recall, increasing photo retrieval effectiveness by over 400% across large-scale media archives.
Categories
Type
AI & Data Science
Need
Benefit
Business function
Industry
Business Model
Business Size
