Text-Guided Alternative Image Clustering
Stephan, Andreas, Miklautz, Lukas, Leiber, Collin, de Araujo, Pedro Henrique Luz, Répás, Dominik, Plant, Claudia, Roth, Benjamin
–arXiv.org Artificial Intelligence
Traditional image clustering techniques only find a single grouping within visual data. In particular, they do not provide a possibility to explicitly define multiple types of clustering. This work explores the potential of large vision-language models to facilitate alternative image clustering. We propose Text-Guided Alternative Image Consensus Clustering (TGAICC), a novel approach that leverages user-specified interests via prompts to guide the discovery of diverse clusterings. To achieve this, it generates a clustering for each prompt, groups them using hierarchical clustering, and then aggregates them using consensus clustering. TGAICC outperforms image- and text-based baselines on four alternative image clustering benchmark datasets. Furthermore, using count-based word statistics, we are able to obtain text-based explanations of the alternative clusterings. In conclusion, our research illustrates how contemporary large vision-language models can transform explanatory data analysis, enabling the generation of insightful, customizable, and diverse image clusterings.
arXiv.org Artificial Intelligence
Jun-7-2024
- Country:
- Oceania > Australia
- North America > United States
- New York > New York County
- New York City (0.04)
- Nebraska > Douglas County
- Omaha (0.04)
- New York > New York County
- Europe
- Austria > Vienna (0.14)
- Monaco (0.04)
- United Kingdom > England
- Greater London > London (0.04)
- Slovenia > Drava
- Municipality of Benedikt > Benedikt (0.04)
- Middle East > Malta
- Eastern Region > Northern Harbour District > St. Julian's (0.04)
- Germany > Bavaria
- Upper Bavaria > Munich (0.04)
- Asia > Middle East
- Jordan (0.04)
- Genre:
- Overview (0.88)
- Research Report (0.84)
- Technology: