Deep learning-based spatiotemporal multi-event reconstruction for delay line detectors: Article in Machine Learning: Science and Technology
Delay line detectors have the advantage of being able to detect the position and time of particle events with high precision while also being multi-hit capable.
Our article presents a novel machine learning approach to enhance the spatiotemporal resolution of delay line detectors for multi-event reconstruction.
By employing deep learning models, we have developed techniques to more accurately detect and reconstruct the position and time of particle events that occur close together, which classical methods struggle with.
Very simplified, the most important point of the correct reconstruction for delay-line detectors is the peak detection of voltage pulses and their matching across different channels. Our neural network allows this even when the peaks are so close that humans or classical electronic hardware cannot distinguish them anymore.