End-to-End Conditional GAN-based Architectures for Image Colourisation

被引:2
|
作者
Blanch, Marc Gorriz [1 ]
Mrak, Marta [1 ]
Smeaton, Alan F. [2 ]
O'Connor, Noel E. [2 ]
机构
[1] BBC Res & Dev, London, England
[2] Dublin City Univ, Dublin, Ireland
基金
欧盟地平线“2020”;
关键词
Colourisation; Conditional GANs; CNNs;
D O I
10.1109/mmsp.2019.8901712
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
In this work recent advances in conditional adversarial networks are investigated to develop an end-to-end architecture based on Convolutional Neural Networks (CNNs) to directly map realistic colours to an input greyscale image. Observing that existing colourisation methods sometimes exhibit a lack of colourfulness, this paper proposes a method to improve colourisation results. In particular, the method uses Generative Adversarial Neural Networks (GANs) and focuses on improvement of training stability to enable better generalisation in large multi-class image datasets. Additionally, the integration of instance and batch normalisation layers in both generator and discriminator is introduced to the popular U-Net architecture, boosting the network capabilities to generalise the style changes of the content. The method has been tested using the ILSVRC 2012 dataset, achieving improved automatic colourisation results compared to other methods based on GANs.
引用
收藏
页数:6
相关论文
共 50 条
  • [21] Fully End-to-End learning based Conditional Boundary Equilibrium GAN with Receptive Field Sizes Enlarged for Single Ultra-High Resolution Image Dehazing
    Ki, Sehwan
    Sim, Hyeonjun
    Choi, Jae-Seok
    Kim, Soo Ye
    Seo, Soomin
    Kim, Saehun
    Kim, Munchurl
    PROCEEDINGS 2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS (CVPRW), 2018, : 930 - 937
  • [22] LLD-GAN: An end-to-end network for low-light image demosaicking
    Wang, Li
    Shi, Cong
    Pundlik, Shrinivas
    Yang, Xu
    Liu, Liyuan
    Luo, Gang
    DISPLAYS, 2024, 85
  • [23] A Survey of End-to-End Driving: Architectures and Training Methods
    Tampuu, Ardi
    Matiisen, Tambet
    Semikin, Maksym
    Fishman, Dmytro
    Muhammad, Naveed
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2022, 33 (04) : 1364 - 1384
  • [24] The MALACH Corpus: Results with End-to-End Architectures and Pretraining
    Picheny, Michael
    Yang, Qin
    Zhang, Daiheng
    Zhang, Lining
    INTERSPEECH 2023, 2023, : 5097 - 5101
  • [25] Towards end-to-end QoS in service oriented architectures
    Ditze, M
    Jahnich, I
    2005 3rd IEEE International Conference on Industrial Informatics (INDIN), 2005, : 92 - 97
  • [26] Attention-based end-to-end image defogging network
    Yang, Yan
    Zhang, Chen
    Jiang, Peipei
    Yue, Hui
    ELECTRONICS LETTERS, 2020, 56 (15) : 759 - +
  • [27] End-to-End Learning-Based Image Compression: A Review
    Chen Jimin
    Lin Zehao
    LASER & OPTOELECTRONICS PROGRESS, 2020, 57 (22)
  • [28] End-to-end image compression method based on perception metric
    Shuai Liu
    Yingcong Huang
    Huoxiang Yang
    Yongsheng Liang
    Wei Liu
    Signal, Image and Video Processing, 2022, 16 : 1803 - 1810
  • [29] A deep learning network based end-to-end image composition
    Zhu, Xiaoyu
    Wang, Haodi
    Zhang, Zhiyi
    Wu, Xiuping
    Guo, Junqi
    Wu, Hao
    SIGNAL PROCESSING-IMAGE COMMUNICATION, 2022, 101
  • [30] A Novel End-to-End Image Caption Based on Multimodal Attention
    Li X.-M.
    Yue G.
    Chen G.-W.
    Dianzi Keji Daxue Xuebao/Journal of the University of Electronic Science and Technology of China, 2020, 49 (06): : 867 - 874