サインイン
インサイト - Visual Computing
Deformable One-shot Face Stylization via DINO Semantic Guidance: A Detailed Analysis
Deformable face stylization framework using DINO semantic guidance enhances one-shot face stylization efficiency.
Advancing Visual Quality Comparison with Co-Instruct Dataset
The author introduces the Co-Instruct dataset to enhance visual quality comparison by providing open-ended settings and detailed reasoning, surpassing existing benchmarks and outperforming state-of-the-art models.
Development of Context-Sensitive Formulas for Constant Luminance Perception
The authors present a framework for developing context-sensitive luminance correction formulas to achieve constant luminance perception for foreground objects by making them slightly translucent. The approach is based on the relative size of the foreground object and polynomial functions.
© 2024 by Linnk AI