cv Download CV

General Information

Full Name Dr.-Ing. Bashir Kazimi
Background Computer Science, Artificial Intelligence
Research Interest and Expertise Deep Learning and Computer Vision
Languages Persian, Turkish, English, and German
Tools/Libraries Python, PyTorch, and Tensorflow

Education

  • 2017-2021
    Ph.D. in Geodesy and Geoinformatics
    Leibniz University Hannover, Hannover, Germany
    • Worked on applications of deep learning in airborne laser scanning data
    • Multiple publications and projects on detection of archaeological objects
    • Dissertation on Self Supervised Learning for Detection of Archaeological Monuments in LiDAR Data supervised by Prof. Dr.-Ing. habil. Monika Sester
  • 2015-2017
    M.Sc. in Artificial Intelligence
    Polytechnic University of Catalonia, Barcelona, Spain
    • Fundamental courses and projects in Machine Learning, Computer Vision, and Natural Language Processing
    • Master Thesis in Neural Machine Translation supervised by Marta Ruiz Costa-jussà
    • Publication at International Journal of the Spanish Society for Natural Language Processing
  • 2010-2015
    B.Sc. in Computer Engineering
    Middle East Technical University (METU), Ankara, Turkey
    • Got a Turkish Government Scholarship for undergraduate studies
    • Was part of a 4-people team that built a social network as a graduation project

Experience

  • Sep. 2023 - Present
    Research Group Leader
    Forschungszentrum Jülich, Institute for Materials Data Science and Informatics (IAS-9)
    • Leading the Data Science and Computer Vision for Electron Microscopy Lab
    • Research on deep learning for electron microscopy data
    • PhD Supervisoin, Deep-Learning Assisted Fast In Situ 4D Electron Microscope Imaging, Helmholtz Imaging Project (From 2024)
  • Feb. 2023 - Aug. 2023
    Postdoctoral Researcher
    Forschungszentrum Jülich, Institute for Materials Data Science and Informatics (IAS-9)
    • Research on deep learning for electron microscopy data
    • Prototyping and implementation of unsupervised, semi‑supervised and supervised learning models
    • Applications of Generative Adversarial Networks (GANs), Autoencoders and Diffusion Models
    • Denoising, super‑resolution and segmentation of transmission electron microscopy (TEM) images of nanoparticles
    • Implementations in PyTorch/Lightning
    • Data pre and post‑processing and visualization with Scikit‑Learn, Pandas, Scikit‑image, and Matplotlib
    • Experiment tracking, monitoring and visualization with Wandb
  • Apr. 2021 - Dec. 2022
    Postdoctoral Researcher
    Helmholtz Zentrum Hereon, Institute of Materials Physics
    • Research on segmentation of bone implants using deep learning and synchrotron radiation computed tomography (CT) data
    • Development of active learning methods well‑suited for tasks with small amounts of annotated data
    • Implementations in PyTorch
    • Data analysis and visualization using Fiji/ImageJ
    • Contribution to a web‑service for the active learning model to help users/domain experts apply segmentation on their data without having to learn or implement deep learning methods
  • Apr. 2017 - Apr. 2021
    Doctoral Researcher
    Leibniz University Hannover, Institute of Cartography and Geoinformatics
    • Research on detection and description of historical man‑made landscape structures
    • Prototyping and implementation of self supervised learning models (GANs and Autoencoders) to leverage large volumes of unlabeled data
    • Transfer learning with self‑supervised pretrained models customized for and finetuned on downstream tasks (classification, semantic and instance segmentation) with limited annotated data
    • Worked on digital terrain models from airborne laser scanning data
    • Worked with the ArcGIS software and Python Osgeo/Gdal library for data processing
    • Used Tensorflow and Keras libraries for implementing deep learning models
    • Published papers and open sourced implementations for classification, semantic segmentation, and instance segmentation of archaeological objects in digital terrain data using deep learning
    • Helped teach master level courses, Internet‑GIS (2017) and Environmental Data Analysis (2018 & 2020)
    • Master Thesis Supervision, Segmentation of Linear Terrain Structures in Digital Terrain Models Using Deep learning by Heyeu Zhang at Leibniz University Hannover (2020)
    • Master Thesis Supervision, Estimation of building parameters from street view images by Xin Hu at Leibniz University Hannover (2020)
    • Master student project supervision, Extraction of linear structures from digital terrain models using deep learning by Ramish Satari at Leibniz University Hannover (2020)
  • Nov. 2017 - Dec. 2017
    Visiting Researcher
    University of Melbourne, Department of Infrastructure Engineering
    • Research visit as part of a scholarship award by DAAD (German Academic Exchange Service)
    • Research collaboration between Institute of Cartography and Geoinformatics at Leibniz University Hannover and Department of Infrastructure Engineering at University of Melbourne
    • Worked on and published a paper for archaeological object detection in airborne laser scanning data
  • Jul. 2016 - Mar. 2017
    Quality Assurance Intern
    Typeform S.L.
    • Automated and manual software tests for features before being shipped for production
  • Feb. 2016 - Jul. 2016
    Java Developer
    Open University of Catalonia
    • Helped improve a website for students to upload programming assignments to be graded automatically

Awards and Fundings

  • 2024
    Helmholtz Foundation Model Initiative (HFMI) Funding 2024
    • Project titled A symbiotic modular foundation model for accelerating solar energy materials development (SOL-AI)
    • Together with KIT, HZB, and Hereon, coordinated by FZJ/IAS-9, we got funded through the HFMI call 2024
  • 2023
    Helmholtz Imaging Project Funding
    • Project titled Deep-Learning Assisted Fast In Situ 4D Electron Microscope Imaging
    • Prof. Stefan Sandfeld and I, together with our collaborators, Prof. Christoph Kirchlechner and Dr. Subin Lee from Institute for Applied Materials – Mechanics of Materials and Interfaces (IAM‑MMI), got the Helmholtz Imaging Project funding of 200K Euros for the proposal of our project planned for 3 years
  • 2022
    Help a hematologist out Challenge
    • 3rd place winning solution
    • I took part in the Help a hematologist out Challenge at Helmholtz Incubator Summer Academy ‑ From Zero to Hero, 2022 and joined the BLAMAD team. The theme of the challenge was to find creative domain adaptation solutions for blood‑cell classification which is important for diagnosis of diseaeses such as anemia or leukemia. We used domain adaptation techniques and won the 3rd place among all participating teams

Skills

Languages Persian (Native), Turkish (Advanced), English (Advanced), German (B2, Goethe Certificate)
Technical Skills Python, PyTorch/Lightning, Tensorflow/Keras, Pandas, Scikit‑Learn, ArcGIS, Osgeo/Gdal, Git, Docker, SQL, Linux, Matploblib, Wandb
Machine Learning Linear/Logistic Regression, Clustering, Convolutional Neural Networks, Classification, Semantic Segmentation, Object Detection, Instance Segmentation, Prototyping, GANs, Autoecnoders, Vision Transformers, Diffusion Models

Presentations/Talks

  • 37th Umbrella Symposium at Forschungszentrum Jülich themed as Advancing energy materials with state-of-the-art analytics and AI, Jülich, Germany, Sep. 2024. Invited talk on Self-Supervised Learning in Electron Microscopy.
  • 17th European Microscopy Congress (EMC), Copenhagen, Denmark, Aug. 2024. Poster presentation on Comparative Analysis of Self-Supervised Learning Techniques for Electron Microscopy Images.
  • IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), Computer Vision for Materials Science Workshop, Seattle, WA, USA, June, 2024. Oral presentation on Self-Supervised Learning with Generative Adversarial Networks for Electron Microscopy.
  • The Eighth Conference on Frontiers of Aberration Corrected Electron Microscopy (PICO24), Vaals, Netherlands, Apr. 2024. Poster presentation on Enhancing Semantic Segmentation in High-Resolution TEM Images: A Comparative Study of Batch Normalization and Instance Normalization.
  • 1st Conference on Artificial Intelligence in Materials Science and Engineering, Saarbrücken, Germany, Nov. 2023. Oral presentation on Enhancing Semantic Segmentation in High‑Resolution TEM Images through Pretraining on Unlabeled Data.
  • 2nd Joint Lab MDMC Workshop, Jülich, Germany, Mar. 2023. Oral presentation on Introduction to the joint ER‑C/IAS‑9 Electron Microscopy Data Science Lab.
  • Helmholtz AI Conference, Dresden, Germany, Jun. 2022. Oral presentation on Deep Active Learning for Segmentation of Biodegradable Bone Implants in High Resolution Synchrotron Radiation Microtomograms.
  • Conference on Cultural Heritage and New Technologies (CHNT 25), Vienna, Austria, Nov. 2020, Oral presentation on Effectiveness of DTM Derivatives for Object Detection Using Deep Learning.
  • Conference on Cultural Heritage and New Technologies (CHNT 24), Vienna, Austria, Nov. 2019. Oral presentation on Semi Supervised Learning for Archaeological Object Detection in Digital Terrain Models.
  • 38. Wissenschaftlich‑Technische Jahrestagung der DGPF e.V., PFGK18, Munich, Germany, Mar. 2018. Poster presentation on Classification of laser scanning data using deep learning.
  • 10th International Conference on Geographic Information Science, GIScience, Melbourne, Australia, Aug. 2018. Oral presentation on Deep learning for archaeological object detection in airborne laser scanning data.
  • The 18th International Conference on Computer Analysis of Images and Patterns, Salerno, Italy, Sep. 2019. Poster presentation on Object instance segmentation in digital terrain models.
  • Joint ISPRS Conference on Photogrammetric Image Analysis and Munich Remote Sensing Symposium, PIA 2019+MRSS, Munich, Germany, Sep. 2020. Oral presentation on Semantic Segmentation of Manmade Landscape Structures in Digital Terrain Models.