Technische Universität Braunschweig
  • Study & Teaching
    • Beginning your Studies
      • Prospective Students
      • Degree Programmes
      • Application
      • Fit4TU
      • Why Braunschweig?
    • During your Studies
      • Fresher's Hub
      • Term Dates
      • Courses
      • Practical Information
      • Beratungsnavi
      • Additional Qualifications
      • Financing and Costs
      • Special Circumstances
      • Health and Well-being
      • Campus life
    • At the End of your Studies
      • Discontinuation and Credentials Certification
      • After graduation
      • Alumni
    • For Teaching Staff
      • Strategy, Offers and Information
      • Learning Management System Stud.IP
    • Contact
      • Study Service Centre
      • Academic Advice Service
      • Student Office
      • Career Service
  • Research
    • Research Profile
      • Core Research Areas
      • Clusters of Excellence at TU Braunschweig
      • Research Projects
      • Research Centres
      • Professors‘ Research Profiles
    • Early Career Researchers
      • Support in the early stages of an academic career
      • PhD-Students
      • Postdocs
      • Junior research group leaders
      • Junior Professorship and Tenure-Track
      • Habilitation
      • Service Offers for Scientists
    • Research Data & Transparency
      • Transparency in Research
      • Research Data
      • Open Access Strategy
      • Digital Research Announcement
    • Research Funding
      • Research Funding Network
      • Research funding
    • Contact
      • Research Services
      • Academy for Graduates
  • International
    • International Students
      • Why Braunschweig?
      • Degree seeking students
      • Exchange Studies
      • TU Braunschweig Summer School
      • Refugees
      • International Student Support
      • International Career Service
    • Going Abroad
      • Studying abroad
      • Internships abroad
      • Teaching and research abroad
      • Working abroad
    • International Researchers
      • Welcome Support for International Researchers
      • Service for Host Institutes
    • Language and intercultural competence training
      • Learning German
      • Learning Foreign Languages
      • Intercultural Communication
    • International Profile
      • Internationalisation
      • International Cooperations
      • Strategic partnerships
      • International networks
    • International House
      • About us
      • Contact & Office Hours
      • News and Events
      • International Days
      • 5th Student Conference: Internationalisation of Higher Education
      • Newsletter, Podcast & Videos
      • Job Advertisements
  • TU Braunschweig
    • Our Profile
      • Aims & Values
      • Regulations and Guidelines
      • Alliances & Partners
      • The University Development Initiative 2030
      • Ecoversity – the TU Braunschweig as a university ecosystem
      • Facts & Figures
      • Our History
    • Career
      • Working at TU Braunschweig
      • Vacancies
    • Economy & Business
      • Entrepreneurship
      • Friends & Supporters
    • General Public
      • Check-in for Students
      • CampusXperience
      • The Student House
      • Access to the University Library
    • Media Services
      • Communications and Press Service
      • Services for media
      • Film and photo permits
      • Advices for scientists
      • Topics and stories
    • Contact
      • General Contact
      • Getting here
  • Organisation
    • Presidency & Administration
      • Executive Board
      • Designated Offices
      • Administration
      • Committees
    • Faculties
      • Carl-Friedrich-Gauß-Fakultät
      • Faculty of Life Sciences
      • Faculty of Architecture, Civil Engineering and Environmental Sciences
      • Faculty of Mechanical Engineering
      • Faculty of Electrical Engineering, Information Technology, Physics
      • Faculty of Humanities and Education
    • Institutes
      • Institutes from A to Z
    • Facilities
      • University Library
      • Gauß-IT-Zentrum
      • Professional and Personnel Development
      • International House
      • The Project House of the TU Braunschweig
      • Transfer Service
      • University Sports Center
      • Facilities from A to Z
    • Equal Opportunity Office
      • Equal Opportunity Office
      • Family
      • Diversity for Students
  • Search
  • Quicklinks
    • People Search
    • Webmail
    • cloud.TU Braunschweig
    • Messenger
    • Cafeteria
    • Courses
    • Stud.IP
    • Library Catalogue
    • IT Services
    • Information Portal (employees)
    • Link Collection
    • DE
    • EN
    • Instagram
    • YouTube
    • LinkedIn
    • Mastodon
    • Bluesky
Menu
  • Organisation
  • Faculties
  • Faculty of Electrical Engineering, Information Technology, Physics
  • Institutes
  • Institute for Communications Technology
  • Research
Logo Institut für Nachrichtentechnik der TU Braunschweig
Computer Vision
  • Computer Vision
    • ⤶ Research
    • ⌂ IfN

Computer Vision

In many research areas, such as automated driving, more and more learned methods based on neural networks are used for visual environment perception. In our group, we investigate semantic and instance segmentation, as well as methods that allow adaptation of the segmentation to unseen data or new classes.

Semantic Segmentation and Instance Segmentation

Semantic segmentation is defined as the pixel-wise classification of a camera image. The goal is to obtain an understanding of the objects present in the image. Using neural networks, each pixel in the image is assigned to one of for example 19 predefined classes. Sometimes, however, this information is not sufficient because overlapping objects of the same class are not distinguished. The solution to this problem is instance segmentation, in which each individual instance of a certain class in the image also receives its own identifier. This makes it possible, for example, to count or track these objects.

Picture with 3 cars on a crossing
Prediction of the objects in a picture

Domain Adaptation for Vision Tasks

Training these aforementioned segmentation networks requires a large amount of labeled data. However, if the data during operation deviates from the training data, for example because they are recorded with a different camera, then the pixel-wise classification may no longer function properly. This is referred to as a so-called domain gap between the data during operation (target data) and the training data. To overcome this domain gap, we develop methods for unsupervised domain adaptation [1] that do not require labels from the intended target domain.

Continual Adaptation and Class-Incremental Learning

Many domain adaptation methods, while not requiring labels of the target domain, depend on training data and (unlabeled) images of the target domain, being available simultaneously during the training and adaptation process. However, this is not always possible, as privacy regulations or storage limitations may prevent the original training data from being made available to the adaptation process. Therefore, our group performs research on methods that allow continual adaptation without training domain data. To achieve this, we are exploring methods such as adapting the network weights [2] or applying a style transfer to the images [3]. Another use case that we are working on is class-incremental learning. This research aims at teaching a network new classes without forgetting the previously learned information [4].

Drei Personen farblich markiert visuelle Erkennung

Publications

[1] J.-A. Bolte, M. Kamp, A. Breuer, S. Homoceanu, P. Schlicht, F. Hüger, D. Lipinski, and T. Fingscheidt, “Unsupervised Domain Adaptation to Improve Image Segmentation Quality Both in the Source and Target Domain,” in Proc. of CVPR - Workshops, Long Beach, CA, USA, Jun. 2019, pp. 1404-1413.

[2] M. Klingner, J.-A. Termöhlen, J. Ritterbach, and T. Fingscheidt, “Unsupervised BatchNorm Adaptation (UBNA): A Domain Adaptation Method for Semantic Segmentation Without Using Source Domain Representations,” arXiv 2011.08502, Nov. 2020, pp. 1-11.

[3] J.-A. Termöhlen, M. Klingner, L. J. Brettin, N. M. Schmidt, and T. Fingscheidt, „Continual Unsupervised Domain Adaptation for Semantic Segmentation by Online Frequency Domain Style Transfer,” in Proc. of ITSC, Indianapolis, IN, USA, Sep. 2021, pp. 1-8.

[4] M. Klingner, A. Bär, P. Donn, and T. Fingscheidt, “Class-Incremental Learning for Semantic Segmentation Re-Using Neither Old Data Nor Old Labels,” in Proc. of ITSC, Rhodes, Greece, Sep. 2020, pp. 1-8.

Photo credits on this page

For All Visitors

Vacancies of TU Braunschweig
Career Service' Job Exchange 
Merchandising

For Students

Term Dates
Courses
Degree Programmes
Information for Freshman
TUCard

Internal Tools

Glossary (GER-EN)
Change your Personal Data

Contact

Technische Universität Braunschweig
Universitätsplatz 2
38106 Braunschweig

P. O. Box: 38092 Braunschweig
GERMANY

Phone: +49 (0) 531 391-0

Getting here

© Technische Universität Braunschweig
Legal Notice Privacy Accessibility

TU Braunschweig uses the software Matomo for anonymised web analysis. The data serve to optimise the web offer.
You can find more information in our data protection declaration.