Explainable Multimedia Retrieval in Mixed Reality (Master Project, Ongoing)
Author
Description
As multimedia retrieval systems evolve to handle increasingly complex queries, their internal decision-making processes often become less transparent. This lack of explainability can hinder user trust and limit the perceived reliability of such systems. Explainability is pivotal in fostering user confidence, empowering users to understand why specific results were retrieved, and enabling them to make informed decisions based on the system's insights.
This project proposes the development of an explainable multimedia retrieval approach. By integrating advanced visualization techniques and using features such as attention maps, metadata associations, or feature similarity metrics, the system will provide users with accessible, real-time explanations for retrieval results. Ultimately, the goal is to enhance user trust, satisfaction, and understanding by making retrieval systems intuitive and transparent. The results should be easily used in a mixed reality environment.
Start / End Dates
2025/02/17 - 2025/06/16