Matthias Körschens

• Deep learning
• Feingranulare Bildklassifikation, -detektion und -segmentierung
• Schwach überwachte Lernverfahren
• Automatische Bildanalyse zur Biodiversitätsforschung
• Auswirkungen des Klimawandels auf Biodiversität, funktionelle Merkmale von Pflanzen und Phänologie

• Seit 10/2022: Wissenschaftliche Mitarbeiter, Flexpool, FSU Jena Institut für Ökologie und Evolution, Professur Biodiversität der Pflanzen
• Seit 04/2019: Doktorand an der FSU Jena, Institut für Ökologie und Evolution, Professur Biodiversität der Pflanzen
• 08/2018 – 03/2019: Wissenschaftliche Hilfskraft der Computer Vision Group an der FSU Jena
• 05/2018: Masterarbeit über das Thema "Identification in Wildlife Monitoring"
• 04/2016 – 05/2018: Masterstudium Informatik an der FSU Jena
• 02/2016: Bachelorarbeit über das Thema "Simulation eines kapazitiven Sensors zur Geometriebestimmung und -bewertung eines Katheters"
• 09/2012 – 02/2016: Bachelorstudium Informatik an der Hochschule Harz in Wernigerode
• 07/2012: Abitur am Domgymnasium Merseburg

Bodesheim, P., Blunk, J., Körschens, M., Brust, C. A., Käding, C., & Denzler, J. (2022). Pre-trained models are not enough: active and lifelong learning is important for long-term visual monitoring of mammals in biodiversity research—Individual identification and attribute prediction with image features from deep neural networks and decoupled decision models applied to elephants and great apes. Mammalian Biology 102, 853–875.

Körschens, M., Bodesheim, P., & Denzler, J. (2022). Beyond Global Average Pooling: Alternative Feature Aggregations for Weakly Supervised Localization. Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP, 180-191. DOI: 10.5220/0010871700003124.

Körschens, M., Bodesheim, P., & Denzler, J. (2022). Occlusion-Robustness of Convolutional Neural Networks via Inverted Cutout. In 2022 26th International Conference on Pattern Recognition (ICPR) (pp. 2829-2835). IEEE.

Gruner, B., Körschens, M., Barz, B., & Denzler, J. (2021). Domain Adaptation and Active Learning for Fine-Grained Recognition in the Field of Biodiversity. arXiv preprint arXiv:2110.11778.

Körschens, M., Bodesheim, P., Römermann, C., Bucher, S.F., Migliavacca, M., Ulrich, J., Denzler, J. (2021) Weakly Supervised Segmentation Pretraining for Plant Cover Prediction. In: Bauckhage C., Gall J., Schwing A. (eds) Pattern Recognition. DAGM GCPR 2021. Lecture Notes in Computer Science, vol 13024. Springer, Cham. https://doi.org/10.1007/978-3-030-92659-5_38

Körschens, M., Bodesheim, P., Römermann, C., Bucher, S.F., Migliavacca, M., Ulrich, J., Denzler, J. (2021) Automatic Plant Cover Estimation with Convolutional Neural Networks, CS4BioDiversity Workshop der INFORMATIK2021

Körschens, M., Bodesheim, P., Römermann, C., Bucher, S.F., Ulrich, J., Denzler, J. (2020) Towards Confirmable Automated Plant Cover Determination. In: Bartoli A., Fusiello A. (eds) Computer Vision – ECCV 2020 Workshops. ECCV 2020. Lecture Notes in Computer Science, vol 12540. Springer, Cham. https://doi.org/10.1007/978-3-030-65414-6_22

Körschens, M., Denzler, J. (2019) ELPephants: A Fine-Grained Dataset for Elephant Re-Identification. The IEEE International Conference on Computer Vision (ICCV) Workshop

Körschens, M., Barz, B., Denzler J. (2018) Towards Automatic Identification of Elephants in the Wild, AI for Wildlife Conservation Workshop (AIWC).

Konferenzbeiträge

Körschens, M., Römermann, C., Bucher, S. F., Ulrich, J., Denzler, J. (2019) Deep Learning Approaches for Automatic Analysis of Plant Species and Cover Determination. iDiv-Konferenz Leipzig & GfÖ Jahrestagung Münster, Poster

Körschens M., Barz B., Denzler, J. (2018) Towards Automatic Identification of Elephants in the Wild. ICEI 2018, Jena, Poster