[IKs] 画像感性 特別講演

[IKs-01] Kansei Imaging: Holistic Image Enhancement

*Carl Staelin¹, Boris Oicherman² (1. Google Inc., 2. University of Minnesota)

We describe and explore how Kansei Imaging should combine and integrate research and algorithms from a wide range of disciplines, such as image analysis, image processing, machine learning, printing, photography, painting and composition, color science, and beauty styling (makeup). We explore how the final application impacts nearly all aspects of the imaging solution: e.g. more user feedback and guidance permits stronger and "riskier" enhancements and alternatives. We also explore how integrating various analyses and segmentations into the solution in order to allow control over each enhancement on a per-pixel basis is vital. We finally raise topics for further research.