The paper introduces the first large-scale crop-composition-explanation benchmark for aesthetic cropping and shows that an SFT+GRPO pipeline trained on it beats prior explainable-cropping methods on IoU and explanation quality.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
COMEX: A Composition-Grounded Benchmark and Learning Framework for Explainable Aesthetic Image Cropping
The paper introduces the first large-scale crop-composition-explanation benchmark for aesthetic cropping and shows that an SFT+GRPO pipeline trained on it beats prior explainable-cropping methods on IoU and explanation quality.