Available Thesis Topics

If you are interested in working on one of the below topics, please contact Dr. Anett Hoppe (anett.hoppe[at]tib.eu.).

In your email, state

• the type of student project that you are interested in (lab, bachelor or master thesis) ,
• the topic(s) your are interested in (please limit yourself to your top 3).

(See the topics in detail)

• Analysis of Student Drawings

• Historical Visual Question Answering 

• Identification of Narrative Patterns in News Videos

• Named Entity Linking in Scientific Videos 

• Multimodal Hate Speech Detection for Videos

• 3D Sports Field Registration (Camera Calibration)  

• Event Detection in News Images

• Geolocation Estimation using Auxiliary Information

• Draft classification in artworks

• Image-text relationship in art literature

• Knowledge graph for art documents

• Pace Analysis of Educational Videos

• Analysis of Informal Language in Educational Videos

• Patents Similarity Ranking using Multimodal Features

• Clustering Patent Image Types

• Temporal Segmentation of Instructional Videos based on Multimodal similarities

• How Videos are changing Human Mood and Personality?

• Categorizing and Linking Equations in Learning Material

• News recommendation using image-text Relations and user characteristics

• Generating task-agnostic data for understanding Vision-Language models

• Incorporating context or auxiliary information in Vision-language models

• Multimodal fake news detection in social media

 

Completed:

• Classification of figures and relevant text relations in scientific documents
  edited by Sebastian Holzki

• Person identification in image data of the internet archive
  edited by Sebastian Diering 

• Aligning and visualising positon and video data of football games

• Automatic classification of team formations in football games using position data
  edited by Jonas Theiner 

• Evaluation of data augmentation techniques for formula recognition in images 

• Content analysis and result visualization for video analytics

• Automatic analysis of student drawings in chemistry lessons using machine learning

• A system for exploring relations between scientific publications and software implementations
  edited by Jascha Hagen
  completed in November 2017 

• Image geolocation estimation using context information
  edited by Kader Pustu-Iren 
  completed in October 2017

• Enriching scientific videos with information from open data sources
  completed in October 2017

• Synthesizing face variations from limited data for face recognition
  completed in May 2017 

• Estimating the information gap between textual and graphical representations
  edited by Christian Henning
  completed in January 2017