Presents Reason50K dataset and ReasonBrain framework for hypothetical instruction-based image editing that requires physical, temporal, causal, and story reasoning.
The objective is defined as: LMLLM =− rX i=1 logp {θ∪θLoRA} ([IMGi]|IA(E I(I)), RV , RT ,E T (H),[IMG 1]
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Reasoning to Edit: Hypothetical Instruction-Based Image Editing with Visual Reasoning
Presents Reason50K dataset and ReasonBrain framework for hypothetical instruction-based image editing that requires physical, temporal, causal, and story reasoning.