In this paper, we present the effectiveness of image compression based on a convolutional auto encoder (CAE) with region of interest (ROI) for quality control. We propose a method that adapts image quality for prioritized parts and non-prioritized parts for CAE-based compression. The proposed method uses annotation information for the distortion weights of the MS-SSIM-based loss function. We show experimental results using a road damage image dataset that is used to check damaged parts and an image dataset with segmentation data (ADE20K). The experimental results reveals that the proposed weighted loss function with CAE-based compression from F. Mentzer et al. learns some characteristics and preferred bit allocations of the prioritized parts by end-to-end training. In the case of using road damage image dataset, our method reduces bpp by 31% compared to the original method while meeting quality requirements that an average weighted MS-SSIM for the road damaged parts be larger than 0.97 and an average weighted MS-SSIM for the other parts be larger than 0.95.
Hiroaki AKUTSU
Hitachi, Ltd.
Takahiro NARUKO
Hitachi, Ltd.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Hiroaki AKUTSU, Takahiro NARUKO, "End-to-End Deep ROI Image Compression" in IEICE TRANSACTIONS on Information,
vol. E103-D, no. 5, pp. 1031-1038, May 2020, doi: 10.1587/transinf.2019EDP7264.
Abstract: In this paper, we present the effectiveness of image compression based on a convolutional auto encoder (CAE) with region of interest (ROI) for quality control. We propose a method that adapts image quality for prioritized parts and non-prioritized parts for CAE-based compression. The proposed method uses annotation information for the distortion weights of the MS-SSIM-based loss function. We show experimental results using a road damage image dataset that is used to check damaged parts and an image dataset with segmentation data (ADE20K). The experimental results reveals that the proposed weighted loss function with CAE-based compression from F. Mentzer et al. learns some characteristics and preferred bit allocations of the prioritized parts by end-to-end training. In the case of using road damage image dataset, our method reduces bpp by 31% compared to the original method while meeting quality requirements that an average weighted MS-SSIM for the road damaged parts be larger than 0.97 and an average weighted MS-SSIM for the other parts be larger than 0.95.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2019EDP7264/_p
Copy
@ARTICLE{e103-d_5_1031,
author={Hiroaki AKUTSU, Takahiro NARUKO, },
journal={IEICE TRANSACTIONS on Information},
title={End-to-End Deep ROI Image Compression},
year={2020},
volume={E103-D},
number={5},
pages={1031-1038},
abstract={In this paper, we present the effectiveness of image compression based on a convolutional auto encoder (CAE) with region of interest (ROI) for quality control. We propose a method that adapts image quality for prioritized parts and non-prioritized parts for CAE-based compression. The proposed method uses annotation information for the distortion weights of the MS-SSIM-based loss function. We show experimental results using a road damage image dataset that is used to check damaged parts and an image dataset with segmentation data (ADE20K). The experimental results reveals that the proposed weighted loss function with CAE-based compression from F. Mentzer et al. learns some characteristics and preferred bit allocations of the prioritized parts by end-to-end training. In the case of using road damage image dataset, our method reduces bpp by 31% compared to the original method while meeting quality requirements that an average weighted MS-SSIM for the road damaged parts be larger than 0.97 and an average weighted MS-SSIM for the other parts be larger than 0.95.},
keywords={},
doi={10.1587/transinf.2019EDP7264},
ISSN={1745-1361},
month={May},}
Copy
TY - JOUR
TI - End-to-End Deep ROI Image Compression
T2 - IEICE TRANSACTIONS on Information
SP - 1031
EP - 1038
AU - Hiroaki AKUTSU
AU - Takahiro NARUKO
PY - 2020
DO - 10.1587/transinf.2019EDP7264
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 5
JA - IEICE TRANSACTIONS on Information
Y1 - May 2020
AB - In this paper, we present the effectiveness of image compression based on a convolutional auto encoder (CAE) with region of interest (ROI) for quality control. We propose a method that adapts image quality for prioritized parts and non-prioritized parts for CAE-based compression. The proposed method uses annotation information for the distortion weights of the MS-SSIM-based loss function. We show experimental results using a road damage image dataset that is used to check damaged parts and an image dataset with segmentation data (ADE20K). The experimental results reveals that the proposed weighted loss function with CAE-based compression from F. Mentzer et al. learns some characteristics and preferred bit allocations of the prioritized parts by end-to-end training. In the case of using road damage image dataset, our method reduces bpp by 31% compared to the original method while meeting quality requirements that an average weighted MS-SSIM for the road damaged parts be larger than 0.97 and an average weighted MS-SSIM for the other parts be larger than 0.95.
ER -