Using Minimum Cost Multicut clustering on foundation-model embeddings, the authors find DINOv2 yields broad visual frames while ConvNeXt V2 yields fine-grained clusters on a climate change image dataset.
Using Multimodal Foundation Models and Clustering for Improved Style Ambiguity Loss
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Teaching text-to-image models to be creative involves using style ambiguity loss, which requires a pretrained classifier. In this work, we explore a new form of the style ambiguity training objective, used to approximate creativity, that does not require training a classifier or even a labeled dataset. We then train a diffusion model to maximize style ambiguity to imbue the diffusion model with creativity and find our new methods improve upon the traditional method, based on automated metrics for human judgment, while still maintaining creativity and novelty.
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
I Spy With My Little Eye: A Minimum Cost Multicut Investigation of Dataset Frames
Using Minimum Cost Multicut clustering on foundation-model embeddings, the authors find DINOv2 yields broad visual frames while ConvNeXt V2 yields fine-grained clusters on a climate change image dataset.