Current options for AI master projects
under supervision of prof. Lambert Schomaker
- Neuromorphic computing (together with prof. Tamalika Banerjee, Materials Science)
- SPICE-level modeling / ensemble of van der Pol oscillators
- Deep learning for microscopic cell analysis in biology (with prof. Ida van der Kleij, Biology)
- Deep learning for classification of truck-load X-ray images (with the Dutch customs)
How to detect weapons and contraband in such images? (the real ones have a much higher resolution,
and there are +50k images).
How to bootstrap such a detector if labeling is extremely sparse?
An excursion to their facility is possible.
The contract entails signing a confidentiality agreement.
- From spectral images of car paint appearance to chemical recipes using deep learning (At a company).
- Time series analysis for industrial maintenance processes (follow-up of the Mantis predictive maintenance project).
Context: Together with PhD student Emmanuel Okafor.
See the Mantis book
- Dead Sea Scrolls (with Maruf Dhali)
- Arabic text recognition using CNN/LSTM / semi-supervised deep learning (with Mahya Ameryan)
- Physics modeling of controls in industrial panels for RL robot simulations of machine tending (with Sha Luo)
- Machine translation and LSTM-based handwriting recognition (with Gideon Maillette)
- Robotics grasping, vision (with Hamidreza Kasaei)
- Project on deep learning for fast conveyor-belt picking at Bollegraaf company.
- Lifelong machine learning / continuous machine learning in historical documents
- Modeling selective attention with spiking neural networks (with Sander Bohte)
Note: This text will never be up to date
Last update: Thu Jan 31 11:47:51 CET 2019