Organizing the "2nd Workshop on Maritime Computer Vision", featuring paper submissions, challenges and keynotes. Find it at https://macvi.org/workshop/macvi24.

Dr. Benjamin Kiefer

Background

Since 2019

PhD student at the Department of Cognitive Systems, University of Tübingen

2016-2019

M. Sc. Mathematics, University of Tübingen

2013 - 2016

B. Sc. Mathematics, University of Tübingen

Research Interests

Currently, I'm working on Deep Learning methods for Object Detection. More specifially, I try to improve Object Detectors by means of freely available meta/environmental data. This meta data helps object detectors be more robust.

Teaching Assistantships

Deep Learning (Winter 2023)
Mobile Robots (Summer 2023)
Artifical Intelligence (Winter 2022)
Introduction to Neural Networks (Summer 2022)
Proseminar: Topics in Deep Neural Networks (Winter 2021)
Team Project: Developing a Web-based Evaluation Application for Deep Neural Network Tasks (Summer 2021)
Einführung in die technische Informatik (Winter 2020)
Mathematik II (Summer 2020)
Deep Neural Networks (Winter 2019)

Various other math courses assisted as a tutor (Numerical Analysis, Stochastics for Computer Scientists, Mathematics for Computer Scientists).

Maritime Computer Vision

Initiated and organized the 1st Workshop on Maritime Computer Vision. Stay tuned for 2nd iteration. Find updates on https://seadronessee.cs.uni-tuebingen.de/ and soon on https://www.macvi.org/.

Avalon Project

I'm part of the Avalon project. This project's goal is to build autonomous vision-based AIs aboard a UAV to assist in maritime emergencies at north sea and baltic sea.

Generating Synthetic Data for UAV scenarios via DeepGTAV

With ever more realistic simulations, we can generate data synthetically to augment or even replace real data entirely. For this reason, jointly with David Ott, we adapted and improved the tool DeepGTAV to work for UAV scenarios. If you are looking for the synthetically generated object detection data for UAV scenarios, find it here. Find the tool here.

SeaDronesSee Benchmark

We created an object detection and tracking benchmark for the use case of maritime search and rescue. Find it here or more information on the corresponding Github repository.

CURRENTLY PAUSED - Neuro Meeting - Paper reading club

The Neuro meeting is a kind of paper reading club, which meets roughly every two weeks to discuss research and applications all kinds of foundations and novelties of neural networks. The central goal of this group is to broaden the horizon and keep an overview over this fast-paced field.

See a list of proposed and past topics at

https://cloud.cs.uni-tuebingen.de/index.php/s/a2gXdMG39cYFJbP

If you are interested to listen to or hold a talk or join regularly just drop me a mail.

Supervised Theses

2021	Bachelor thesis	Accelerating DNN Training Using Weight Extrapolation
2021	Bachelor thesis	Maritime Anomaly Detection Using Autoencoders
2020	Bachelor thesis	Domain Adaptation for Object Detection on UAVs
2020	Bachelor thesis	Sampling Useful Synthetic Data for UAV Object Detection

Publications

[1]	Benjamin Kiefer, Timon Höfer, Andreas Zell; Stable Yaw Estimation of Boats from the Viewpoint of UAVs and USVs (2023, accepted for publication at ECMR 2023) [arxiv]
[2]	Benjamin Kiefer, Yitong Quan, Andreas Zell; Memory Maps for Video Object Detection and Tracking on UAVs (2023, accepted for publication at IROS 2023) [arxiv]
[3]	Pablo Ruiz-Ponce, David Ortiz-Perez, Jose Garcia-Rodriguez, Benjamin Kiefer; POSEIDON: A Data Augmentation Tool for Small Object Detection Datasets in Maritime Environments. In Sensors 23.7 (2023) [Sensors]
[4]	Benjamin Kiefer, et al. 1st Workshop on Maritime Computer Vision (MaCVi) 2023: Challenge Results. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) Workshops, 2023 [arxiv, cvf]
[5]	Timon Höfer, Benjamin Kiefer, Martin Messmer, Andreas Zell; HyperPosePDF - Hypernetworks Predicting the Probability Distribution on SO(3). In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2023, pp. 2369-2379 [cvf]
[6]	Benjamin Kiefer, Andreas Zell. "Fast Region of Interest Proposals on Maritime UAVs" (2022, accepted for publication at ICRA 2023) [arxiv]
[7]	Benjamin Kiefer, David Ott, and Andreas Zell. Leveraging synthetic data in object detection on unmanned aerial vehicles. In 2022 26th International Conference on Pattern Recognition (ICPR), pages 3564--3571, 2022. (*equal contribution) [ieee, arxiv]
[8]	Leon Amadeus Varga, Benjamin Kiefer, Martin Messmer, and Andreas Zell. SeaDronesSee: A Maritime Benchmark for Detecting Humans in Open Water. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 2260--2270, January 2022. (equal contribution) [ieee, cvf, arxiv]
[9]	Martin Messmer, Benjamin Kiefer, and Andreas Zell. Gaining scale invariance in uav bird’s eye view object detection by adaptive resizing. In 2022 26th International Conference on Pattern Recognition (ICPR), pages 3588--3594, 2022 (*equal contribution) [ieee, arxiv]
[10]	Benjamin Kiefer, Martin Messmer, and Andreas Zell. Diminishing Domain Bias by Leveraging Domain Labels in Object Detection on UAVs. In 2021 20th International Conference on Advanced Robotics (ICAR), pages 523--530, December 2021. (*equal contribution) [ieee, arxiv]

Reviews

Reviewed for:

ICIP 2022
ICPR 2022
ECCV 2022
WACV 2023
ICRA 2023
CVPR 2023