Identifeye: Federation of Image Recognition Services

August 30, 2016

In the fast-paced digital landscape, enterprises relied heavily on efficient image recognition services for numerous tasks—ranging from streamlining workflows to enhancing customer experiences. With several powerful computer vision service providers, such as Blippar, CamFind, Wolfram Language, Clarifai, Microsoft Cognitive Services, and TinEye, organizations faced the challenge of integrating and managing these systems for optimized performance. To address this, we developed an MVP of a federated layer for enterprise image recognition queries.

Federating Computer Vision Services

This federated layer acted as an intelligent intermediary, allowing enterprises to query multiple computer vision services either simultaneously or selectively. Instead of being tied to a single provider, businesses were able to leverage the strengths of different platforms for specific tasks—one might have excelled at object recognition, while another was ideal for facial analysis or reverse image searches.

Platform, Testing, and QA Tools

The MVP was designed to be a flexible and scalable platform. It included:

Core Platform:

The core system enabled seamless API integration with top image recognition service providers like Blippar, Clarifai, and Microsoft Cognitive Services.

Testing Tools:

Automated tools benchmarked each service provider's performance across various use cases and image types, offering enterprises a clear understanding of which service to use for specific needs.

Quality Assurance (QA) Tools:

Tools monitored service consistency and accuracy, helping enterprises maintain high standards in image recognition workflows.

Flexibility and Scalability

We adopted a subscription-based business model, where enterprises paid for access to the federated layer and its associated testing and QA tools. The system saved time, reduced integration complexity, and improved the overall efficiency of image recognition workflows.

Key Benefits:

  • Unified API access to multiple computer vision services
  • Automated performance benchmarking
  • Real-time service quality monitoring
  • Cost optimization through intelligent service selection
  • Scalable architecture supporting growing enterprise needs

Technical Implementation

Architecture:

  • RESTful API gateway for service integration
  • Load balancing for optimal performance
  • Caching layer for frequently requested analyses
  • Monitoring and analytics dashboard

Integration Features:

  • Standardized API responses across services
  • Automatic failover and redundancy
  • Custom rules engine for service selection
  • Comprehensive usage analytics

Future Development

Our roadmap included several exciting enhancements:

  • Integration with emerging computer vision services
  • Advanced machine learning for optimal service selection
  • Enhanced analytics and reporting capabilities
  • Expanded testing and benchmarking tools
  • Custom model training and deployment options