The search functionality is under construction.
The search functionality is under construction.

Hierarchical Detailed Intermediate Supervision for Image-to-Image Translation

Jianbo WANG, Haozhi HUANG, Li SHEN, Xuan WANG, Toshihiko YAMASAKI

  • Full Text Views

    0

  • Cite this

Summary :

The image-to-image translation aims to learn a mapping between the source and target domains. For improving visual quality, the majority of previous works adopt multi-stage techniques to refine coarse results in a progressive manner. In this work, we present a novel approach for generating plausible details by only introducing a group of intermediate supervisions without cascading multiple stages. Specifically, we propose a Laplacian Pyramid Transformation Generative Adversarial Network (LapTransGAN) to simultaneously transform components in different frequencies from the source domain to the target domain within only one stage. Hierarchical perceptual and gradient penalization are utilized for learning consistent semantic structures and details at each pyramid level. The proposed model is evaluated based on various metrics, including the similarity in feature maps, reconstruction quality, segmentation accuracy, similarity in details, and qualitative appearances. Our experiments show that LapTransGAN can achieve a much better quantitative performance than both the supervised pix2pix model and the unsupervised CycleGAN model. Comprehensive ablation experiments are conducted to study the contribution of each component.

Publication
IEICE TRANSACTIONS on Information Vol.E106-D No.12 pp.2085-2096
Publication Date
2023/12/01
Publicized
2023/09/14
Online ISSN
1745-1361
DOI
10.1587/transinf.2023EDP7025
Type of Manuscript
PAPER
Category
Image Processing and Video Processing

Authors

Jianbo WANG
  The University of Tokyo
Haozhi HUANG
  Tencent AI Lab
Li SHEN
  Tencent AI Lab
Xuan WANG
  Tencent AI Lab
Toshihiko YAMASAKI
  The University of Tokyo

Keyword