Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis

Research output: Working paper › Preprint › Research

Standard

Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis. / Shen, Flora Yu; Luo, Katie; Yang, Guandao; Haraldsson, Harald; Belongie, Serge.

arXiv.org, 2022.

Research output: Working paper › Preprint › Research

Harvard

Shen, FY, Luo, K, Yang, G, Haraldsson, H & Belongie, S 2022 'Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis' arXiv.org. <https://arxiv.org/pdf/2202.04036.pdf>

APA

Shen, F. Y., Luo, K., Yang, G., Haraldsson, H., & Belongie, S. (2022). Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis. arXiv.org. https://arxiv.org/pdf/2202.04036.pdf

Vancouver

Shen FY, Luo K, Yang G, Haraldsson H, Belongie S. Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis. arXiv.org. 2022.

Author

Shen, Flora Yu ; Luo, Katie ; Yang, Guandao ; Haraldsson, Harald ; Belongie, Serge. / Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis. arXiv.org, 2022.

Bibtex

@techreport{6c9ef60219ce4e439375c5f939c50463,
title = "Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis",
abstract = "In this work, we address an important problem of optical see through (OST) augmented reality: non-negative image synthesis. Most of the image generation methods fail under this condition, since they assume full control over each pixel and cannot create darker pixels by adding light. In order to solve the non-negative image generation problem in AR image synthesis, prior works have attempted to utilize optical illusion to simulate human vision but fail to preserve lightness constancy well under situations such as high dynamic range. In our paper, we instead propose a method that is able to preserve lightness constancy at a local level, thus capturing high frequency details. Compared with existing work, our method shows strong performance in image-to-image translation tasks, particularly in scenarios such as large scale images, high resolution images, and high dynamic range image transfer.",
author = "Shen, {Flora Yu} and Katie Luo and Guandao Yang and Harald Haraldsson and Serge Belongie",
year = "2022",
language = "English",
publisher = "arXiv.org",
type = "WorkingPaper",
institution = "arXiv.org",

}

RIS

TY - UNPB

T1 - Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis

AU - Shen, Flora Yu

AU - Luo, Katie

AU - Yang, Guandao

AU - Haraldsson, Harald

AU - Belongie, Serge

PY - 2022

Y1 - 2022

N2 - In this work, we address an important problem of optical see through (OST) augmented reality: non-negative image synthesis. Most of the image generation methods fail under this condition, since they assume full control over each pixel and cannot create darker pixels by adding light. In order to solve the non-negative image generation problem in AR image synthesis, prior works have attempted to utilize optical illusion to simulate human vision but fail to preserve lightness constancy well under situations such as high dynamic range. In our paper, we instead propose a method that is able to preserve lightness constancy at a local level, thus capturing high frequency details. Compared with existing work, our method shows strong performance in image-to-image translation tasks, particularly in scenarios such as large scale images, high resolution images, and high dynamic range image transfer.

AB - In this work, we address an important problem of optical see through (OST) augmented reality: non-negative image synthesis. Most of the image generation methods fail under this condition, since they assume full control over each pixel and cannot create darker pixels by adding light. In order to solve the non-negative image generation problem in AR image synthesis, prior works have attempted to utilize optical illusion to simulate human vision but fail to preserve lightness constancy well under situations such as high dynamic range. In our paper, we instead propose a method that is able to preserve lightness constancy at a local level, thus capturing high frequency details. Compared with existing work, our method shows strong performance in image-to-image translation tasks, particularly in scenarios such as large scale images, high resolution images, and high dynamic range image transfer.

UR - https://arxiv.org/abs/2202.04036

M3 - Preprint

BT - Residual Aligned: Gradient Optimization for Non-Negative Image Synthesis

PB - arXiv.org

ER -

ID: 303686285