Close

Dr.-Ing. Sarah Blum

Researcher, Software Developer

About Me

I am a newly-graduated post-doc researching advanced signal processing methods for time series data. Currently, I am working in a neuropsychology work group where I investigate anomaly detection methods for data recorded from the brain in everyday situations. My main research interest lies in the advancement of signal processing systems which can detect and decode brain signals using mobile EEG hardware. Currently, I am using Information Geometry methods which aim to apply for example Riemannian geometry directly to the EEG data in order to classify or statistically investigate them. In my second job as software developer at Hörtech I develop tools for hearing aid research. Currently, we are working on a virtual environment to simulate different auditory landscapes.

Experience

Project Leader for Software Development

Hörtech AG, Oldenburg

Managing a Cross-Platform development project aiming to develop desktop and mobile applications to help people to hear better.

Post Doc, University of Oldenburg

Researching on signal processing methods for time series data.

PhD Student

University of Oldenburg

Advances in Signal Processing and Hardware Development for EEG-based Brain-Computer Interfaces

Software Developer and Consultant

Hörtech AG, Oldenburg

Development of hearing aid signal processing algorithms and frameworks to incorporate EEG data into the processing pipeline.

Projects

SCALA

In the field of neuroergonomics, everything is about gaining insights in the human brain and its workings during natural motion, outside of traditional laboratory settings. We added to this field by developing an Android framework for the online processing of EEG and other time series data. SCALA uses the LabStreamingLayer to simultaneously process different data types on standard smartphones. It can be used for Brain Computer Interface experiments and is an open source project.

View Project

Riemannian Artifact Subspace Reconstruction

A big part of working on mobile data recordings are methods for the detection of artifacts in the data. In a recent project, I investigate whether the online artifact correction method Artifact Subspace Reconstruction can be improved by using Riemannian geometry in its core computation steps. Spoiler alert: it can indeed. The Matlab toolbox is an open source project.

View Project

Collection of scientific posters and papers

I use a github repository as a server for my own publications and contributions to conferences.

View Project

Java Artifact Subspace Reconstruction

We are currently working on an integration of ASR into SCALA on Android smartphones. This integration would allow for an online data correction during mobile measurements. More Infos follow here soon.

View Project

The fEEGrid: new flex-printed EEG sensor hardware

In our latest paper, we report on the development and evaluation of the fEEGrid, a new sensor headband for long-term EEG recordings. It has been developed specifically for use in care sitations and sleep studies. Check out the paper here!

View Project

Interview about my research (only in German)

In a recent interview, I talked about my research. I explain why I work on cognitive aspects of hearing and why I originally wanted to work in dementia research. I describe that I am currently developing new approaches for truly 'smart hearing systems' of the future, so that the devices can infer the true intent of the users.

View Talk (YouTube)

Skills