40-Issue 4

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/2633074

Browse

Browsing 40-Issue 4 by Subject "Image processing"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Deep Portrait Lighting Enhancement with 3D Guidance
    (The Eurographics Association and John Wiley & Sons Ltd., 2021) Han, Fangzhou; Wang, Can; Du, Hao; Liao, Jing; Bousseau, Adrien and McGuire, Morgan
    Despite recent breakthroughs in deep learning methods for image lighting enhancement, they are inferior when applied to portraits because 3D facial information is ignored in their models. To address this, we present a novel deep learning framework for portrait lighting enhancement based on 3D facial guidance. Our framework consists of two stages. In the first stage, corrected lighting parameters are predicted by a network from the input bad lighting image, with the assistance of a 3D morphable model and a differentiable renderer. Given the predicted lighting parameter, the differentiable renderer renders a face image with corrected shading and texture, which serves as the 3D guidance for learning image lighting enhancement in the second stage. To better exploit the long-range correlations between the input and the guidance, in the second stage, we design an imageto- image translation network with a novel transformer architecture, which automatically produces a lighting-enhanced result. Experimental results on the FFHQ dataset and in-the-wild images show that the proposed method outperforms state-of-the-art methods in terms of both quantitative metrics and visual quality.
  • Thumbnail Image
    Item
    PosterChild: Blend-Aware Artistic Posterization
    (The Eurographics Association and John Wiley & Sons Ltd., 2021) Chao, Cheng-Kang; Singh, Karan; Gingold, Yotam; Bousseau, Adrien and McGuire, Morgan
    Posterization is an artistic effect which converts continuous images into regions of constant color with smooth boundaries, often with an artistically recolored palette. Artistic posterization is extremely time-consuming and tedious. We introduce a blend-aware algorithm for generating posterized images with palette-based control for artistic recoloring. Our algorithm automatically extracts a palette and then uses multi-label optimization to find blended-color regions in terms of that palette. We smooth boundaries away from image details with frequency-guided median filtering. We evaluate our algorithm with a comparative user study and showcase its ability to produce compelling posterizations of a variety of inputs. Our parameters provide artistic control and enable cohesive, real-time recoloring after posterization pre-processing.