Text Embedding is Not All You Need: Attention Control for Text-to-Image Semantic Alignment with Text Self-Attention Maps

Documenti analoghi

• Constructing Concept-based Models to Mitigate Spurious Correlations with Minimal Human Effort
  di: Kim, Jeeyung, et al.
  Pubblicazione: (2024)
• Test-Time Alignment of Text-to-Image Diffusion Models via Null-Text Embedding Optimisation
  di: Kim, Taehoon, et al.
  Pubblicazione: (2025)
• Semantic Token Reweighting for Interpretable and Controllable Text Embeddings in CLIP
  di: Kim, Eunji, et al.
  Pubblicazione: (2024)
• Not All Attention Heads Are What You Need: Refining CLIP's Image Representation with Attention Ablation
  di: Lin, Feng, et al.
  Pubblicazione: (2025)
• BiasMap: Leveraging Cross-Attentions to Discover and Mitigate Hidden Social Biases in Text-to-Image Generation
  di: Chakraborty, Rajatsubhra, et al.
  Pubblicazione: (2025)