이미지 복원을 위한 네트워크 파라미터의 동적 업데이트를 위한 기법

  • Published : 2020.04.30

Abstract

최근 많은 연구 결과물에서 빅데이터를 이용하여 학습된 뉴럴 네트워크가 영상 내 노이즈를 제거하는데 매우 효과적임이 입증되었다. 여기에서 한 걸음 더 나아가, 입력으로 주어진 노이즈가 있는 영상의 특징을 분석하여, 사전에 학습된 네트워크의 파라미터를 테스트 타임에 동적으로 업데이트함으로써 주어진 입력 영상을 더욱 잘 처리할 수 있도록 하는 연구들이 시도되고 있다. 본 원고에서는 이와 같이 테스트 타임에 주어지는 입력 영상을 네트워크 학습에 사용하는(self-supervision) 이미지 복원 기법들을 소개한다. 다음으로, 기존의 self-supervision을 이용하는 기법들 대비 학습 효율성과 정확도를 더욱 향상시킬 수 있는 새로운 형태의 네트워크 파라미터 업데이트 기법을 설명하고, 제안하는 기법의 우수성을 다양한 실험 결과를 통해 분석 및 입증한다.

Keywords

References

  1. Dabov, K., Foi, A., Katkovnik, V., Egiazarian, K.: Image denoising by sparse 3-d transform-domain collaborative filtering. IEEE Transactions on Image Processing 16(8), 2080-2095 (2007) https://doi.org/10.1109/TIP.2007.901238
  2. Portilla, J., Strela, V., Wainwright, M.J., Simoncelli, E.P.: Image denoising using scale mixtures of gaussians in the wavelet domain. IEEE Transactions on Image Processing 12(11), 1338-1351 (2003) https://doi.org/10.1109/TIP.2003.818640
  3. Buades, A., Coll, B., Morel, J..: A non-local algorithm for image denoising. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2005)
  4. Ehret, T., Davy, A., Morel, J.M., Facciolo, G., Arias, P.: Model-blind video denoising via frame-to-frame training. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2019)
  5. Krull, A., Buchholz, T.O., Jug, F.: Noise2void-learning denoising from single noisy images. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2019)
  6. Lehtinen, J., Munkberg, J., Hasselgren, J., Laine, S., Karras, T., Aittala, M., Aila, T.: Noise2Noise: Learning image restoration without clean data. In: International Conference on Machine Learning (ICML). vol. 80, pp. 2965-2974 (2018)
  7. Batson, J., Royer, L.: Noise2self: Blind denoising by self-supervision. In: International Conference on Machine Learning (ICML) (2019)
  8. Zhang, K., Zuo, W., Chen, Y., Meng, D., Zhang, L.: Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising. IEEE Transactions on Image Processing 26, 3142-3155 (2017) https://doi.org/10.1109/TIP.2017.2662206
  9. Anwar, S., Barnes, N.: Real image denoising with feature attention. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV) (2019)
  10. Glasner, D., Bagon, S., Irani, M.: Super-resolution from a single image. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV) (2009)
  11. Shocher, A., Cohen, N., Irani, M.: "zero-shot" super-resolution using deep internal learning. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  12. Huang, J.B., Singh, A., Ahuja, N.: Single image super-resolution from transformed self-exemplars. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2015)
  13. Lefkimmiatis, S.: Non-local color image denoising with convolutional neural networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2016)
  14. Liu, D., Wen, B., Fan, Y., Loy, C.C., Huang, T.S.: Non-local recurrent network for image restoration. In: Advances in Neural Information Processing Systems (NIPS) (2018)
  15. Anwar,S.,Barnes,N.:Realimagedenoisingwithfeatureattention. In:Proceedings of the IEEE International Conference on Computer Vision (ICCV) (2019)
  16. Zhang, Y., Tian, Y., Kong, Y., Zhong, B., Fu, Y.: Residual dense network for image restoration. CoRR abs/1812.10477 (2018)
  17. Zhang, K., Zuo, W., Zhang, L.: Ffdnet: Toward a fast and flexible solution for cnn based image denoising. IEEE Transactions on Image Processing 27, 4608-4622 (2018) https://doi.org/10.1109/TIP.2018.2839891
  18. Lee, S., Cho, D., Kim, J., Kim, T.H., Restore from Restored: Single Image Denoising with Pseudo Clean Image. arXiv:2003.04721