Perception Simulation Machine Learning Internship

What you will do: 

We are seeking a talented and driven Intern to join our team in Stuttgart and support our efforts to investigate a data driven failure analysis model that allows us to cluster and trace back errors to scene attributes. To this end, in this internship you will investigate Vision Language Models (VLMs) to tag scenes based on images automatically and use those tags to trace back to performance errors. This tagging will not only help us as a Perception V&V team but is also a viable pseudo labeling approach for our machine learning teams.

Candidates are expected to work 40 hours per week for the duration of the internship.  

They must be willing to travel to or be in Stuttgart, Germany for the duration of the internship. The position is hybrid (partially working from office and partially working from home).  

Torc Internships provide a fantastic way for you to hone your skills and see what life at Torc is all about. Our goal is to enable interns (m/f/d) to become full-time Torc’rs.

Responsibilities:

•  Work with the Perception V&V team to develop a suitable VLM approach to facilitate the event tagging from images
• Implement and fine-tune the models using internal datasets and generate results
• Analyze results by comparing to manually labeled data

What you'll need to succeed: 

• Full-time student (m/f/d) pursuing an undergraduate or graduate degree in computer science, computer engineering or related field 
• Knowledge of Linux platforms 
• Knowledge of dashboarding tools and libraries 
• Software development experience 
• Familiarity with one or more high-level languages: Python, C/C++ 
• Enthusiasm to take on new challenges and learn new skills  
• Willingness to work in an international team 

Bonus Points! 

• Familiarity with Cloud providers (AWS, GCP, Azure) 
• Experience in self-driving car industry 
• Experience with simulations