About Me
Hi, my name is Tim-Oliver and I am a research software engineer. I work on biomedical image processing and analysis pipelines. Next to my work as software engineer I educate researchers in the best practices of FAIR image processing and analysis. I have worked with a number of different research groups and have experience in a wide range of image processing and analysis tasks.
In my day-to-day work I use a variety of different programming languages and tools. I am experienced in Python and Java, and have worked with PyTorch, TensorFlow, scikit-image and Dask. I am used to work with git and continues integration to ensure the quality of the solutions I provide.
I am experienced in working with different imaging modalities from light to electron microscopy and handling all types of imaging data, from small 2D images to large 5D datasets.
I am passionate about making state of the art technology available to people without a computer science background.
Projects
A template to get your image processing and analysis project started.
Image Processing and Analysis (IPA), transforming raw image data into quantitative measurements, is an important part of life sciences. This template simplifies the initiation of new IPA projects and streamlines the reorganization of existing ones, contributing to research that is more Findable, Accessible, Interoperable, and Reusable (FAIR) promoting efficiency, reproducibility, and reliability in life sciences research.
Inspect the point spread function (PSF) of your microscope.
This plugin is used to analyze the point spread function of microscopes. It provides a simple and easy to use interface, which allows us to quickly assess the state of our microscopes. The quality measurements are are collected and stored in a database, which allows us to track the performance of our microscopes over time.
Experience
Machine Learning Expert
Facility for Advanced Imaging and Microscopy
July 2021 - Present
Friedrich Miescher Institute for Biomedical Research
In my role as Machine Learning Expert in the Facility for Advanced Imaging and Microscopy I work together with our research groups to develop and implement state of the art image processing and analysis pipelines. The projects I work on are a diverse mix, ranging from small scale image analysis tasks to processing of terabyte sized imaging datasets. I also educate our users in the best practices of FAIR image processing and analysis.
Doctoral Researcher
Center for Systems Biology Dresden
2017 - 2021
Next to researching novel approaches for image restoration, it was always a passion of mine to make our research outputs accessible to the wider non-computer scientist community. Making state of the art technology available to researchers without a computer science background is a challenge I enjoy.
Student Assistant
Junior Programmer
2014 - 2017
During my bachelor and master studies I worked as student assistant at the Chair for Bioinformatics and Information Mining. During this time I developed KNIME workflows and plugins for image processing and analysis. As part of my work I also contributed to ImageJ and Fiji, two widely used image processing tools in the life sciences.
Education
Technische Universität Dresden
Dr. rer. nat., Computer Science
2017 - 2021
I developed novel machine learning methods for image restoration in the field of biomedical image analysis. As a member of the IMPRS-CellDevoSys school, hosted at the Max Planck Institute of Molecular Cell Biology and Genetics (MPI-CBG) in collaboration with the Technische Universität Dresden (TUD), I was part of a vibrant and interdisciplinary research community. I worked together with researchers from biology, chemistry, physics and computer science.
Universität Konstanz
M. Sc., Computer Science
2016 - 2017
After completing my four year bachelor studies in computer science at the University of Konstanz I enrolled in the one year master program. During my master studies I focused on the basics of computer vision and image processing.
Selected Publications
N2V2 - Fixing Noise2Void Checkerboard Artifacts with Modified Sampling Strategies and a Tweaked Network Architecture
E Höck*, TO Buchholz*, A Brachmann, F Jug, A Freytag
Noise2Void 2 (N2V2) fixes the checkerboard artifacts that are present in some Noise2Void (N2V) denoised images.
cryoCARE - Content-Aware Image Restoration for Cryo-Transmission Electron Microscopy Data
TO Buchholz, M Jordan, G Pigino, F Jug
Even highest quality cryo-EM data is inherently noisy and suffers from low contrast. With cryoCARE we present a deep learning-based approach to denoise and improve contrast in cryo-EM data.
Noise2Void - Learning Denoising from Single Noisy Images
A Krull*, TO Buchholz*, F Jug
Noise2Void is a self-supervised image denoising method based on deep learning. It does not require clean ground truth data for training, instead it can be trained on single noisy images only.
Teaching & Extracurricular Activities
Alongside my daily work I am involved in different teaching activities:
- ZIDAS 2024: Co-organizer of the 2024 edition of the ZIDAS summer school.
- ZIDAS 2023: Co-organizer of the 2023 edition of the ZIDAS summer school.
- ZIDAS 2022: Co-organizer of the 2022 edition of the ZIDAS summer school.
- 2019 Dresden Senioren Akademie: Wie können lernende Algorithmen unsichtbare Inhalte in mikroskopischen Bildern sichtbar machen?
- 2019 Output DD: Computer Vision Meets Biology - Basic Research Requires Open Source Solutions