Dr. Nikolaus Mayer Computer Vision Lab Department of Computer Science University of Freiburg Germany Georges-Köhler-Allee, Building 052 79110 Freiburg E-Mail:

Brief Bio	Scientific Interests	Publications	Projects	Teaching	Miscellaneous

---

---

Brief Bio

Nikolaus Mayer received his Bachelor's degree in Computer Science in 2011 and his Master's degree in Applied Computer Science in 2014 at the University of Freiburg, Germany. Each time he was honored as best of the year.

For his Master's thesis work he was awarded the 2014 VDI-Förderpreis (sponsorship award of the Association of German Engineers).

In 2014 he joined the Computer Vision Group in Freiburg under Prof. Thomas Brox as a full-time PhD student, and went on to graduate in 2020.

---

Scientific Interests

• Neural networks for 3D vision
• Synthetic data generation for deep learning

Input	Output
	Sorry, this display widget needs JavaScript.
	(Viewer requires WebGL support)

---

Publications

Master's Thesis

• N. Mayer, "Coupling ICP and Whole Image Aligment for Real-time Camera Tracking", Albert-Ludwigs-Universität Freiburg, January 2014 [pdf]

---

Projects

TrimBot2020

• The TrimBot2020 project is funded by the ERC and brings together partners from all over Europe

• We are building an autonomous gardening robot, with lots of vision and robotics components working hand-in-hand

---

Teaching

Summer term 2019

• Seminar: Current Works in Computer Vision (Paper topic: Unsupervised learning of correspondence)

Winter term 2018/2019

• Seminar: Current Works in Computer Vision (Paper topic: Deep Object Pose Estimation)

Summer term 2016

• Seminar: Current Works in Computer Vision (Paper topic: Initialization of Neural Networks)

Winter term 2015/2016

• Lecture assistance: Image Processing and Computer Graphics
• Seminar: Current Works in Computer Vision (Paper topic: View synthesis using ConvNets)

Summer term 2015

• Lecture assistance: Statistical Pattern Recognition

Winter term 2014/2015

• Lecture assistance: Image Processing and Computer Graphics
• Seminar: Current Works in Computer Vision (Paper topic: 3D reconstruction of nonrigid objects)
• Proseminar: Bildverarbeitung mit ImageJ und OpenCV (Topics: Scanline Disparity Estimation, Rolling Shutter Artifacts)

Summer term 2014

• Lecture assistance: Computer Vision II
• Proseminar: Bildverarbeitung mit ImageJ und OpenCV (Topic: Scanline Disparity Estimation)

Student's theses and projects supervision

• Bachelor's thesis: "Enabling and Analyzing Mixed Precision Training in Convolutional Neural Networks", 2020

• Master's thesis: "A CNN Based Blind Approach to Lossy Image Compression Artefact Reduction", 2020

• Master's thesis: "Quantifying and improving CNN robustness against common image degradations", 2019 (together with Huizhong Zhou)

• Master project: "Joint SfM and disparity network for depth estimation in stereo videos", 2018 (together with Benjamin Ummenhofer)

• Bachelor's thesis: "Evaluation of Dataset Characteristics for Deep Network Training", 2017

• Master project: "Automated Anthropometric Measurements with a Consumer Depth Camera", 2015

---

Miscellaneous

Some talks and code

• CMake is hard. ROS is harder. People seem to like my CMake+ROS project templates.

• Hands-on course on building a raytracing engine from scratch.

• Introductory presentation about (open-source) software licenses.

• High-level introductory presentation about C++11.

• ImageJ plugin: Z Profiler 2, a z-axis profile plotter that correctly deals with hyperstacks.

  Download the source file and follow the plugin installation guide.

---

---