Teledyne FLIR, a subsidiary of Teledyne Technologies Incorporated, a major player in the development of infrared imaging, intelligent systems and detection solutions, announces the addition of a new solution to its Teledyne FLIR Prism software family: Prism AIMMGen.
Prism AIMMGen is an ITAR-exempt AI model generation service that enables the automated creation of AI and machine learning models using synthetically generated data. It allows system integrators developing AI/ML products for commercial, first response or defense applications to achieve significant savings in terms of costs and time.
An answer to the challenges of collecting real-world data
Developing AI/ML models traditionally requires manual data collection and labeling, time-consuming and costly tasks, particularly in contexts where the objects to be detected are rare or difficult to capture in real-world scenarios, such as military targets. Prism AIMMGen overcomes these obstacles by generating millions of ML-ready examples of synthetic objects under varying conditions and across different spectrums, including visible light and infrared. This process drastically reduces engineering time and costs, allowing high-quality models to be produced in days rather than weeks. Its internal MLOps infrastructure ensures that data is secure and models can be developed, deployed and maintained efficiently.
Dan Walker, Vice President of Teledyne FLIR, explains:
“With the pace and sophistication of systems development, it is essential to leverage intelligent synthetic data to train best-in-class AI models. AIMMGen accelerates time-to-market for developers while ensuring their respective AI products are ready for all possible scenarios in virtually any environmental conditions.”
Prism AIMMGen Key Features
- Unlimited generation of synthetic images : Access to a vast data lake not subject to ITAR restrictions, applicable to multiple use cases in object detection.
- Specific datasets : Ability to produce millions of automatically annotated synthetic images, adapted to the specific needs of various applications, with an interactive search tool for optimized learning.
- Automated model optimization : Rapid generation of high-performance AI models with a secure MLOps infrastructure, making it possible to meet the needs of developers within very short deadlines.
Follow the reference Newsletter on artificial intelligence (more than 18,000 members), daily and 100% free.
