Master Thesis Open-World 3D Anomaly Segmentation with Vision-Language Models (f/m/x)

BMW M Motorsport / BMW Group; München; Vor Ort; EUR 60.000 - 80.000; Vollzeit

Zusammenfassung

A leading company offers a Master thesis opportunity focusing on 3D anomaly segmentation using vision-language models. This project aims to leverage innovative approaches for 3D environments, providing students an excellent chance for personal and professional development in collaboration with a reputable academic institution.

Leistungen

Comprehensive mentoring & onboarding

Flexible working hours

Digital offers & mobile working

Attractive remuneration

Apartment offers for students

Qualifikationen

• Students with a strong background in deep learning and computer vision.
• Ideally be enrolled at TUM.
• Prior experience with vision-language models or semantic segmentation is a plus.

Aufgaben

• Develop methods for 3D semantic segmentation from existing annotations.
• Evaluate 2D segmentation foundations for 3D segmentation.
• Compare feature representations for improved segmentation.

Kenntnisse

Deep Learning

Computer Vision

3D Scene Understanding

Anomaly Detection

Ausbildung

Currently enrolled in Master's program

Background in computer science or related field

Jobbeschreibung

We, the BMW Group, offer you an interesting and varied Master thesisin the field of open-world 3D anomaly segmentation, focusing on leveraging contrastive vision-language models to understand and segment unknown objects in 3D environments. This project will utilize the Open World 3D dataset, which provides both 3D bounding boxes and 2D image data, enabling innovative multi-modal segmentation approaches.

What awaits you?

• Vision-language models for open-world perception.
• 3D segmentation using bounding box annotations.
• Pseudo-labeling techniques with segmentation models.
• Anomaly detection through contrastive feature representations.
• Develop novel methods to generate 3D semantic segmentation data from existing annotations.
• Evaluate open-vocabulary 2D segmentation as a foundation for 3D segmentation.
• Compare various feature representations, including BEV-based heatmaps, for improved segmentation.

This thesis will be supervised by a professor at Technical University of Munich (TUM). Please note that your thesis must be supervised by a university on your part.

What should you bring along?

• Students with a strong background in deep learning and computer vision.
• Ideally be enrolled at TUM.
• 3D scene understanding.
• Prior experience with vision-language models,multi-view projectionor semantic segmentation is a plus but not mandatory.

If you are passionate about advancing open-world 3D anomaly segmentation and working with cutting-edge vision-language models, we look forward to your application!

What do we offer?

• Comprehensive mentoring & onboarding.
• Personal & professional development.
• Flexible working hours.
• Digital offers & mobile working.
• Attractive remuneration.
• Apartment offers for students (subject to availability & only Munich).
• And many other benefits - seebmw.jobs/benefits

Earliest starting date: from19.05.2025

Duration: 6 months

Working hours:Fulltime

Do you have any questions? Then simply send your enquiry using ourcontact form. Your enquiry will then be answered by telephone or e-mail.

At the BMW Group, we see diversity and inclusion in all its dimensions as a strength for our teams. Equal opportunities are a particular concern for us, and the equal treatment of applicants and employees is a fundamental principle of our corporate policy. That is why our recruiting decisions are also based on personality, experience and skills.

Find out more about diversity at the BMW Group atbmwgroup.jobs/diversity