Repository logo
Home
Research Outputs
Statistics
  1. Home
  2. SuUB
  3. Dissertationen
  4. The role of visual grounding in visual question answering generalization and shortcut learning
Citation link
10.26092/elib/3092

The role of visual grounding in visual question answering generalization and shortcut learning

Date Issued
2024-06-06
Author(s)
Reich, Daniel
Supervisors
Schultz, Tanja
Putze, Felix
Referees
Stiefelhagen, Rainer
Schultz, Tanja
Abstract
We set out on this thesis’ journey with a conceptual idea of what Visual Grounding’s role and impact in VQA should be, but soon learn that - apart from the interference of shortcut learning with these expectations - there are also other fundamental limitations of contemporary Visual Grounding research that impede an in-depth analysis, namely a lack of well-defined procedures to measure Visual Grounding in VQA and properly evaluate its impact, We address these limitations with our contributions and in the process gain a much clearer understanding about why the impact of Visual Grounding on VQA performance has been difficult to grasp and how its role can be better highlighted with appropriately designed evaluation scenarios. Finally, in the last chapter, this thesis culminates in the definition of a theoretical model (VGR) that clearly describes the role of Visual Grounding in the context of VQA generalization and shortcut learning, thereby marking the end of our journey.
Subjects
Visual Question Answering

; 

Visual Grounding

; 

Shortcut Learning
Institution
Universität Bremen
Faculty
Fachbereich 03: Mathematik/Informatik (FB 03)
Type
Dissertation
License
https://creativecommons.org/licenses/by/4.0/
Language
English
File(s)
Loading...
Thumbnail Image
Download
Name

Dissertation_Daniel_Reich_The_Role_of_VG_in_VQA_Generalization_and_Shortcut_Learning.pdf

Description
Dissertation Daniel Reich: The Role of VG in VQA Generalization and Shortcut Learning
Size

10.12 MB

Format

Adobe PDF

Checksum

(MD5):93c54ed1d41166d52f7c81f28d161473

Built with DSpace-CRIS software - Extension maintained and optimized by 4Science

  • Accessibility settings
  • Privacy policy
  • End User Agreement
  • Send Feedback