B.Sc./M.Sc. Thesis Offer: "Obstacle Detection on Lunar Surfaces with Deployment for Real-Time Processing"
Motivation:
We are offering an exciting opportunity for a Bachelors/Master's thesis student to
work on an advanced computer vision project focused on lunar surface analysis. The
goal is to develop and benchmark instance segmentation models for identifying and
segmenting rocks and craters on the lunar surface. This SAMLER-KI project will
combine state-of-the-art machine learning techniques with real-world applications in
planetary exploration.
Key Responsibilities:
•Conduct a comprehensive literature review of the latest instance segmentation
models and techniques.
•Train instance segmentation models on diverse datasets, including real, synthetic,
and mixed lunar data.
•Evaluate and benchmark the performance of trained models using appropriate
metrics.
•Optimize and deploy the best-performing model on an FPGA platform for real-time
processing.
•Document research findings, methodology, and results for academic and industrial
dissemination.
Prior Knowledge:
•Strong knowledge of machine learning and deep learning techniques, particularly in
computer vision.
•Hands-on experience with frameworks such as TensorFlow, PyTorch, or similar.
•Proficiency in training and fine-tuning instance segmentation models (e.g., Mask RCNN,
YOLO, or transformers).
•Familiarity with dataset preparation, augmentation, and handling mixed data
sources.
•Knowledge of hardware deployment on FPGA or similar platforms (preferred).
•Programming skills in Python and hardware-specific languages (e.g.,OpenCL for
FPGA)(optional)
We look forward to receiving your complete informative application documents
(please include CV, Cover letter, transcript of grades and certificates). or further
information and application, please contact Siddhant Shete, siddhant.shete[at]dfki.de
Deutsches Forschungszentrum für
Künstliche Intelligenz GmbH (DFKI)
Robotics Innovation Center
Siddhant Shere
Robert-Hooke-Str. 1
28359 Bremen
www.dfki.de/ric
Phone: +49 421 17845 4179
siddhant.shete[at]dfki.de