Source author record

Stefan Meier

Stefan Meier appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Catalog footprint

What is connected

2works
3topics
3close collaborators

Actions

Connect this record

Log in to claim

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

2 published item(s)

preprint2021arXiv

Coulomb interactions and the spatial coherence of femtosecond nanometric electron pulses

The transverse coherence of electrons is of utmost importance in high resolution electron microscopes, point-projection microscopes, low-energy electron microscopy and various other applications. Pulsed versions of many of these have recently been realized, mostly relying on femtosecond laser-triggering electron emission from a sharp needle source. We here observe electron interference fringes and measure how the interference visibility becomes reduced as we increase the electron bunch charge. Due to the extremely strong spatio-temporal confinement of the electrons generated here, we observe the visibility reduction already at average electron bunch charges of less than 1 electron per pulse, owing to the stochastic nature of the emission process. We can fully and quantitatively explain the loss of coherence based on model simulations. Via the van Cittert-Zernike theorem we can connect the visibility reduction to an increase of the effective source size. We conclude by discussing emittance, brightness and quantum degeneracy, which have direct ramifications to many setups and devices relying on pulsed coherent electrons.

preprint2016arXiv

Efficient Target Activity Detection based on Recurrent Neural Networks

This paper addresses the problem of Target Activity Detection (TAD) for binaural listening devices. TAD denotes the problem of robustly detecting the activity of a target speaker in a harsh acoustic environment, which comprises interfering speakers and noise (cocktail party scenario). In previous work, it has been shown that employing a Feed-forward Neural Network (FNN) for detecting the target speaker activity is a promising approach to combine the advantage of different TAD features (used as network inputs). In this contribution, we exploit a larger context window for TAD and compare the performance of FNNs and Recurrent Neural Networks (RNNs) with an explicit focus on small network topologies as desirable for embedded acoustic signal processing systems. More specifically, the investigations include a comparison between three different types of RNNs, namely plain RNNs, Long Short-Term Memories, and Gated Recurrent Units. The results indicate that all versions of RNNs outperform FNNs for the task of TAD.