IMEC has developed GpflowOpt, a novel Python framework for Bayesian optimization, which can be viewed as a modern spin-off of the widely used SUMO-toolbox. GpflowOpt allows the user to speed up expensive simulations or the tuning of deep learning vision systems. In the latter case, it achieves this by replacing such as grid-search by optimized sequential parameter tuning, drastically reducing the training time of such systems while achieving better or equivalent performance. The software has been adopted to allow more accurate paper-cardboard separation by among others more efficient tuning of deep learning vision technology and the ability to process diverse compositions of paper-cardboard of varying quality by rapid reconfiguration.

Area of the technology

  • Machine Learning
  • Surrogate modelling
  • Bayesian Optimization
  • Hyperparameter tuning of machine learning systems

Targeted Industrial Sectors

  • Engineering
  • 3D printing
  • Automotive
  • Recycling industry

Technology Readiness level

TRL 6 - technology demonstrated in relevant environment

Contact Information

Dirk Deschrijver