This survey provides a taxonomy of preference alignment methods for diffusion models in image generation and editing, covering optimization techniques, applications, and open challenges.
Drivinggpt: Unifying driving world modeling and planning with multi-modal autoregressive transformers
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Preference Alignment on Diffusion Model: A Comprehensive Survey for Image Generation and Editing
This survey provides a taxonomy of preference alignment methods for diffusion models in image generation and editing, covering optimization techniques, applications, and open challenges.