access icon free Accurate single image super-resolution using cascading dense connections

© The Institution of Engineering and Technology
Received 11/03/2019, Published 21/05/2019

In recent years, deep learning methods, especially deep convolutional neural network, have been successfully applied to the single-image super-resolution (SISR) task. In this Letter, the authors propose an accurate SISR method by introducing cascading dense connections in a very deep network. In detail, they construct the cascading dense network (CDN) to fully make use of the hierarchical features from all the convolutional layers, which implements a cascading mechanism upon the dense connected convolutional layers. The cascading dense connection in the CDN enables short and long paths to be built directly from the output to each layer, alleviating the vanishing-gradient problem of very deep networks. Extensive experiments show that CDN achieves state-of-the-art performance on traditional metrics (PSNR and SSIM). Also, they introduce object recognition as the additional evaluation metric for SISR, which further demonstrates the effectiveness of the authors’ method.

Inspec keywords: convolution; neural nets; learning (artificial intelligence); object recognition; image resolution

Other keywords: dense connected convolutional layers; deep learning methods; accurate single image super-resolution; single-image super-resolution task; deep convolutional neural network; CDN; accurate SISR method; cascading mechanism; cascading dense network; cascading dense connection; deep network

Subjects: Optical, image and video signal processing; Neural computing techniques; Computer vision and image processing techniques; Knowledge engineering techniques

References

    1. 1)
      • 11. Zhang, Y., Tian, Y., Kong, Y., et al: ‘Residual dense network for image super-resolution’. Proc. Computer Vision Pattern Recognition (CVPR), Salt Lake City, Utah, USA, 2018, pp. 2472–2481, doi: 10.1109/CVPR.2018.00262.
    2. 2)
      • 8. Tai, Y., Yang, J., Liu, X., et al: ‘Memnet: a persistent memory network for image restoration’. Proc. IEEE Int. Conf. on Computer Vision (ICCV), Venice, Italy, 2017, pp. 4539–4547, doi: 10.1109/ICCV.2017.486.
    3. 3)
      • 22. Agustsson, E., Timofte, R.: ‘NTIRE 2017 challenge on single image super-resolution: dataset and study’. Proc. Computer Vision Pattern Recognition Workshops (CVPRW), Honolulu, Hawaii, USA, 21 July 2017, pp. 1121–1131, doi: 10.1109/CVPRW.2017.150.
    4. 4)
      • 9. Lai, W.S., Huang, J.B., Ahuja, N., et al: ‘Deep laplacian pyramid networks for fast and accurate super-resolution’. Proc. Computer Vision Pattern Recognition (CVPR), Honolulu, Hawaii, USA, 2017, pp. 624–632, doi: 10.1109/CVPR.2017.618.
    5. 5)
      • 17. Haris, M., Shakhnarovich, G., Ukita, N.: ‘Deep backprojection networks for super-resolution’. Proc. Computer Vision Pattern Recognition (CVPR), Salt Lake City, Utah, USA, 2018, pp. 1664–1673, doi: 10.1109/CVPR.2018.00179.
    6. 6)
      • 19. Ren, H., EI-Khamy, M., Lee, J.: ‘Image super resolution based on fusing multiple convolution neural networks’. Proc. Computer Vision Pattern Recognition Workshops (CVPRW), Honolulu, Hawaii, USA, 2017, pp. 1050–1057, doi: 10.1109/CVPRW.2017.142.
    7. 7)
      • 12. Shi, W., Caballero, J., Huszar, F., et al: ‘Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network’. Proc. Computer Vision Pattern Recognition (CVPR), Las Vegas, Nevada, USA, 2016, pp. 1874–1883, doi: 10.1109/CVPR.2016.207.
    8. 8)
      • 4. Lim, B., Son, S., Kim, H., et al: ‘Enhanced deep residual networks for single image super-resolution’. Proc. Computer Vision Pattern Recognition Workshops (CVPRW), Honolulu, Hawaii, USA, 2017, pp. 136–144, doi: 10.1109/CVPRW.2017.151.
    9. 9)
      • 6. Zhang, Y., Li, K., Li, K., et al: ‘Image super-resolution using very deep residual channel attention networks’. Proc. European Conf. Computer Vision (ECCV), Munich, Germany, 2018, pp. 294–310, doi: 10.1007/978-3-030-01234-2_18.
    10. 10)
    11. 11)
      • 13. Dong, C., Loy, C.C., Tang, X.: ‘Accelerating the super-resolution convolutional neural network’. Proc. European Conf. Computer Vision (ECCV), Amsterdam, Netherlands, 2016, pp. 391–407, doi: 10.1007/978-3-319-46475-6_25.
    12. 12)
      • 16. Kim, J., Kwon Lee, J., Mu Lee, K.: ‘Accurate image super-resolution using very deep convolutional networks’. Proc. Computer Vision Pattern Recognition (CVPR), Las Vegas, Nevada, USA, 2016, pp. 1646–1654, doi: 10.1109/CVPR.2016.182.
    13. 13)
      • 18. Li, J., Fang, F., Mei, K., et al: ‘Multi-scale residual network for image super-resolution’. Proc. European Conf. Computer Vision (ECCV), Munich, Germany, 2018, pp. 527–542, doi: 10.1007/978-3-030-01237-3_32.
    14. 14)
      • 15. He, K., Zhang, X., Ren, S., et al: ‘Deep residual learning for image recognition’. Proc. Computer Vision Pattern Recognition (CVPR), Las Vegas, Nevada, USA, 2016, pp. 770–778, doi: 10.1109/CVPR.2016.90.
    15. 15)
    16. 16)
      • 21. Hui, Z., Wang, X., Gao, X.: ‘Fast and accurate single image super-resolution via information distillation network’. Proc. Computer Vision Pattern Recognition (CVPR), Salt Lake City, Utah, USA, 2018, pp. 723–731, doi: 10.1109/CVPR.2018.00082.
    17. 17)
      • 14. Szegedy, C., Liu, W., Jia, Y., et al: ‘Going deeper with convolutions’. Proc. Computer Vision Pattern Recognition (CVPR), Boston, Massachusetts, USA, 2015, pp. 1–9, doi: 10.1109/CVPR.2015.7298594.
    18. 18)
      • 5. Tai, Y., Yang, J., Liu, X.: ‘Image super-resolution via deep recursive residual network’. Proc. Computer Vision Pattern Recognition (CVPR), Honolulu, Hawaii, USA, 2017, pp. 3147–3155, doi: 10.1109/CVPR.2017.298.
    19. 19)
      • 10. Ahn, N., Kang, B., Sohn, K.A.: ‘Fast, accurate, and, lightweight super-resolution with cascading residual network’. Proc. European Conf. Computer Vision (ECCV), Munich, Germany, 2018, pp. 256–272, doi: 10.1007/978-3-030-01249-6_16.
    20. 20)
      • 7. Kim, J., Kwon Lee, J., Mu Lee, K.: ‘Deeply-recursive convolutional network for image super-resolution’. Proc. Computer Vision Pattern Recognition (CVPR), Las Vegas, Nevada, USA, 2016, pp. 1637–1645, doi: 10.1109/CVPR.2016.181.
    21. 21)
      • 20. Lai, W.S., Huang, J.B., Ahuja, N., et al: ‘Fast and accurate image super-resolution with deep laplacian pyramid networks’, Trans. Pattern Anal. Mach. Intell., 2018, doi: 10.1109/TPAMI.2018.2865304.
    22. 22)
http://iet.metastore.ingenta.com/content/journals/10.1049/el.2019.0392
Loading

Related content

content/journals/10.1049/el.2019.0392
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Loading