Dr. Frank Schindler

Projektmitarbeiter

Kontakt
Telefon: +49 228 73-54197
Fax: +49 228 73-4571
E-Mail: schindl@REMOVETHISPART.cs.uni-bonn.de
   
Besucheradresse: Institut für Informatik 4
Friedrich-Hirzebruch-Allee 8
D-53115 Bonn
Germany
Postanschrift: Friedrich-Hirzebruch-Allee 5
Raum: 1.010
 

Betreute Arbeiten

Sommersemester 2024

  • Analysing the use of foundation models for tracking in wildlife videos (Masterarbeit)

Wintersemester 2023/24

  • Tracking von Wildtieren in Videos mit Hilfe von Pseudo-Depth Informationen  (Bachelorarbeit)
  • Diffusion Models for Tracking and Instance Segmentation of Wildlife Animals (Bachelorarbeit)
  • Semantic Segmentation und Instance Segmentation mit HRNet (Projektgruppe)

Sommersemester 2023

  • Transfer Learning für Instance Segmentation für Tag- und Nachtaufnahmen von Wildtieren mit Hilfe von Centermask (Projektgruppe)
  • Optimierung eines interaktiven Annotationstools für die Nutzung von Videodaten (Projektgruppe)

 

Wintersemester 2022/23

  • Comparison of Supervised and Unsupervised Tracking of Animals in Wildlife Monitoring (Bachelorarbeit)
  • Instance Segmentation mit RDSNet (Projektgruppe)

 

Sommersemester 2022

  • Tracking Objects as Points (Projektgruppe)

Wintersemester 2021/22

  • Refinement von detektierten Instanzmasken in Videos von Wildtieren durch die Betrachtung von Objektmerkmalen im Detail- und Grenzbereich (Bachelorarbeit)
  • The surprising impact of mask-head architecture on novel class segmentation (Projektgruppe)
  • Sipmask: Spatial information preservation for fast image and video instance segmentation (Projektgruppe)

Sommersemester 2021

  • Variational Autoencoders for Instance Segmentation in Wildlife Crossing Videos (Bachelorarbeit)
  • On Mask Propagation for Video Instance Segmentation (Bachelorarbeit)
  • Top-DB-Net: Top DropBlock for Activation Enhancement in Person Re-Identification (Projektgruppe)
  • Lightweight Multi-Branch Network for Person Re-Identification (Projektgruppe)

Wintersemester 2020/21

  • Recurrent metric networks and batch multiple hypothesis for multi-object tracking (Projektgruppe)
  • Customized Multi-Person Tracker (Projektgruppe)

Publikationen

Action detection for wildlife monitoring with camera traps based on segmentation with filtering of tracklets (SWIFT) and mask-guided action recognition (MAROON)

Schindler, F., Steinhage, V., van Beeck Calkoen, S. and Heurich, M. (2024, January). In Applied Sciences 14(2):514.

SWIFT - an efficient and effective application of instance segmentation and tracking in wildlife monitoring

Schindler, F. and Steinhage, V. (2023, September). In Workshop Online Proceedings "Camera Traps, AI and Ecology", Jena.

Instance segmentation and tracking of animals in wildlife videos: SWIFT-segmentation with filtering of tracklets

Schindler, F., and Steinhage, V. (2022, November). In Ecological Informatics 71: 101794.

Saving costs for video data annotation in wildlife monitoring

Schindler, F., and Steinhage, V. (2021, November). In Ecological Informatics 65: 101418.

Identification of animals and recognition of their actions in wildlife videos using deep learning techniques

Schindler, F., and Steinhage, V. (2021, March). In Ecological Informatics 61: 101215.

Semantic labeling and reconstruction of grape bunches from 3D range data using a new RGB-D feature descriptor

Mack, J., Schindler, F. , Rist, F., Herzog, K., Topfer, R., and Steinhage, V. (2018, December). In Computers and Electronics in Agriculture 155: 96-102.