Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Regional Style and Color Transfer

Zhicheng Ding, Panfeng Li, Qikai Yang, Siyang Li, Qingtian Gong

TL;DR

A new approach that leverages a segmentation network to precisely isolate foreground objects within the input image and applies style transfer exclusively to the background region is proposed, resulting in a visually unified and aesthetically pleasing final composition.

Abstract

This paper presents a novel contribution to the field of regional style transfer. Existing methods often suffer from the drawback of applying style homogeneously across the entire image, leading to stylistic inconsistencies or foreground object twisted when applied to image with foreground elements such as person figures. To address this limitation, we propose a new approach that leverages a segmentation network to precisely isolate foreground objects within the input image. Subsequently, style transfer is applied exclusively to the background region. The isolated foreground objects are then carefully reintegrated into the style-transferred background. To enhance the visual coherence between foreground and background, a color transfer step is employed on the foreground elements prior to their rein-corporation. Finally, we utilize feathering techniques to achieve a seamless amalgamation of foreground and background, resulting in a visually unified and aesthetically pleasing final composition. Extensive evaluations demonstrate that our proposed approach yields significantly more natural stylistic transformations compared to conventional methods.

Regional Style and Color Transfer

TL;DR

A new approach that leverages a segmentation network to precisely isolate foreground objects within the input image and applies style transfer exclusively to the background region is proposed, resulting in a visually unified and aesthetically pleasing final composition.

Abstract

This paper presents a novel contribution to the field of regional style transfer. Existing methods often suffer from the drawback of applying style homogeneously across the entire image, leading to stylistic inconsistencies or foreground object twisted when applied to image with foreground elements such as person figures. To address this limitation, we propose a new approach that leverages a segmentation network to precisely isolate foreground objects within the input image. Subsequently, style transfer is applied exclusively to the background region. The isolated foreground objects are then carefully reintegrated into the style-transferred background. To enhance the visual coherence between foreground and background, a color transfer step is employed on the foreground elements prior to their rein-corporation. Finally, we utilize feathering techniques to achieve a seamless amalgamation of foreground and background, resulting in a visually unified and aesthetically pleasing final composition. Extensive evaluations demonstrate that our proposed approach yields significantly more natural stylistic transformations compared to conventional methods.
Paper Structure (12 sections, 7 equations, 4 figures)

This paper contains 12 sections, 7 equations, 4 figures.

Table of Contents

  1. Introduction
  2. Methods
  3. Segmentation and Boundary Optimization
  4. Artistic Style Transfer
  5. Color Transfer
  6. Blending Style and Color
  7. Results
  8. Segmentation and Boundary Optimization
  9. Background Style Transfer
  10. Foreground Color Transfer
  11. Image Blending
  12. Conclusions

Figures (4)

  • Figure 1: Regional style and color transfer framework
  • Figure 2: Comparison of Segmentation Boundary with or without Optimization
  • Figure 3: showcases the effects of global style transfer (Fig \ref{['fig:global_style_transfer']}, background style transfer (Fig \ref{['fig:background_style_transfer']}), foreground color transfer (Fig \ref{['fig:foreground_color_transfer']}) and final blend image (Fig \ref{['fig:blend']}).
  • Figure 4: illustrates a qualitative comparison of the results obtained for six sets of input image (a) and their corresponding style image (b). The comparison includes four images manipulation techniques: global style transfer (c), background-only style transfer (d), foreground-only color transfer (e), and our proposed method – regional style and color transfer (f), which combines background style transfer with foreground color transfer.