Working Student – Machine Learning, Computer Vision

Posted 14hrs ago

Employment Information

Education
Salary
Experience
Job Type

Report this job

Job expired or something wrong with this job?

Job Description

Werkstudent entwickelt Few-Shot-Computer-Vision zur kamerabasierten Speisenerkennung. menio baut modulare Software für die Betriebsgastronomie auf europäischen und internationalen Märkten.

Responsibilities:

  • Research and prototype-compare current approaches to few-shot recognition and visual similarity search
  • Investigate DINOv2, SigLIP, and lightweight models such as MobileNet
  • Use publicly available food datasets for pretraining and benchmarking
  • Build a well-documented dataset together using images from real-world cafeterias
  • Develop a method for adding new dishes from 1–5 reference images without complete retraining
  • Identify unknown dishes using open-set recognition
  • Improve robustness to variations in portion sizes, lighting, and presentation
  • Account for domain shifts between different cafeterias
  • Combine few-shot recognition with detection or segmentation to identify multiple dishes individually on a tray
  • Investigate how the daily menu and SigLIP’s image–text capabilities can improve recognition
  • Define metrics and test scenarios, and analyze confusion cases involving similar dishes
  • Work on a camera-based food recognition system for contract catering

Requirements:

  • Currently pursuing a degree in Computer Science, Data Science, Mathematics, or a comparable field
  • Personal projects using Python, for example as part of your studies, privately, or on GitHub
  • Fundamental knowledge of machine learning and deep learning
  • Interest in computer vision
  • Experience with PyTorch is a plus
  • Initial experience with image classification, embeddings, visual similarity search, or object detection is a plus, but not required
  • Basic knowledge of or familiarity with few-shot learning, transfer learning, open-set recognition, or segmentation
  • Willingness to try new approaches and learn about new topics
  • Application documents submitted as a single PDF, including a cover letter, résumé, and certificates

Benefits:

  • Flexible working hours and remote work
  • Choice of part-time or full-time employment
  • Free lunch three times a week
  • Open work culture, short decision-making processes, and agile working methods
  • Plenty of opportunity to make an impact
  • Modern office facilities in Hanover or Plauen
  • Fresh fruit, excellent coffee or tea, and a selection of beverages tailored to each location
  • Regular team events and outings
  • Bike leasing
  • Corporate benefits
  • Company pension scheme and financial support for occupational retirement provision
  • Numerous professional development and training opportunities
  • Attendance at international conferences
  • Top-floor lounge with rooftop terrace