TY - GEN
T1 - Deep Structured Energy-Based Image Inpainting
AU - Altinel, Fazil
AU - Ozay, Mete
AU - Okatani, Takayuki
N1 - Funding Information:
ACKNOWLEDGEMENTS This work was partly supported by CREST, JST Grant Number JPMJCR14D1, and the ImPACT Program Tough Robotics Challenge of the Council for Science, Technology, and Innovation (Cabinet Office, Government of Japan).
Publisher Copyright:
© 2018 IEEE.
PY - 2018/11/26
Y1 - 2018/11/26
N2 - In this paper, we propose a structured image inpainting method employing an energy based model. In order to learn structural relationship between patterns observed in images and missing regions of the images, we employ an energy-based structured prediction method. The structural relationship is learned by minimizing an energy function which is defined by a simple convolutional neural network. The experimental results on various benchmark datasets show that our proposed method significantly outperforms the state-of-the-art methods which use Generative Adversarial Networks (GANs). We obtained 497.35 mean squared error (MSE) on the Olivetti face dataset compared to 833.0 MSE provided by the state-of-the-art method. Moreover, we obtained 28.4 dB peak signal to noise ratio (PSNR) on the SVHN dataset and 23.53 dB on the CelebA dataset, compared to 22.3 dB and 21.3 dB, provided by the state-of-the-art methods, respectively. The code is publicly available. 1 1 https:llgithub.com/cvlab-tohoku/DSEBImageInpainting.
AB - In this paper, we propose a structured image inpainting method employing an energy based model. In order to learn structural relationship between patterns observed in images and missing regions of the images, we employ an energy-based structured prediction method. The structural relationship is learned by minimizing an energy function which is defined by a simple convolutional neural network. The experimental results on various benchmark datasets show that our proposed method significantly outperforms the state-of-the-art methods which use Generative Adversarial Networks (GANs). We obtained 497.35 mean squared error (MSE) on the Olivetti face dataset compared to 833.0 MSE provided by the state-of-the-art method. Moreover, we obtained 28.4 dB peak signal to noise ratio (PSNR) on the SVHN dataset and 23.53 dB on the CelebA dataset, compared to 22.3 dB and 21.3 dB, provided by the state-of-the-art methods, respectively. The code is publicly available. 1 1 https:llgithub.com/cvlab-tohoku/DSEBImageInpainting.
UR - http://www.scopus.com/inward/record.url?scp=85059749943&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85059749943&partnerID=8YFLogxK
U2 - 10.1109/ICPR.2018.8546025
DO - 10.1109/ICPR.2018.8546025
M3 - Conference contribution
AN - SCOPUS:85059749943
T3 - Proceedings - International Conference on Pattern Recognition
SP - 423
EP - 428
BT - 2018 24th International Conference on Pattern Recognition, ICPR 2018
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 24th International Conference on Pattern Recognition, ICPR 2018
Y2 - 20 August 2018 through 24 August 2018
ER -