Vol. 1 No. 1 (2022): July
RESEARCH ARTICLES

Weakly supervised learning for raindrop removal on an image

DOWNLOAD PDF
  • Abhishek Papanur,
  • Aditya Hebbar,
  • Akash H Hirur,
  • Alok Bhusanur
Abhishek Papanur
School of Computer Science and Engineering, REVA University, Bengaluru, India
Aditya Hebbar
School of Computer Science and Engineering, REVA University, Bengaluru, India
Akash H Hirur
School of Computer Science and Engineering, REVA University, Bengaluru, India
Alok Bhusanur
School of Computer Science and Engineering, REVA University, Bengaluru, India

Published 2022-08-14

Keywords

  • Rain drop removal,
  • machine learning,
  • training rain datasets,
  • raindrop detection,
  • weakly supervised learning, attention

How to Cite

Abhishek Papanur, Aditya Hebbar, Akash H Hirur, & Alok Bhusanur. (2022). Weakly supervised learning for raindrop removal on an image. International Journal of Computational Learning & Intelligence, 1(1), 25–30. Retrieved from http://www.milestoneresearch.in/JOURNALS/index.php/IJCLI/article/view/21

Copyright (c) 2022 Abhishek Papanur, Aditya Hebbar, Akash H Hirur, Alok Bhusanur

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Abstract

 In this research, we address the challenge of removing view-disturbing raindrops from a single picture. Machine learning-based methods show promising for addressing this problem, but they require comprehensive paired photographs for training, i.e., the raindrop-degraded image and the comparable clean image of the same scenario. In the lack of pairing training examples, we propose a weakly supervised learning-based model that requires simply a collection of photos with image-level annotations indicating the presence/absence of raindrops for training. In a multi-task learning approach, we train the raindrops detector to highlight raindrop regions. Following that, we present an attention-based generative network for raindrop removal, as well as a weighted preservation loss for retaining non-raindrop information. Our model, in particular, may be mixed and trained using pairs and unpaired samples, allowing us to easily adapt the model to a new domain. The Experiment validates the effectiveness of the proposed technique. Using only weakly supervised learning, our technique was able to obtain comparable outcomes to heavily supervised learning methods.

DOWNLOAD PDF

References

  1. Roser, M., & Geiger, A. (2009, September). Video-based raindrop detection for improved image registration. In 2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops (pp. 570-577). IEEE.
  2. Yamashita, A., Kuramoto, M., Kaneko, T., & Miura, K. T. (2003, October). A virtual wiper-restoration of deteriorated images by using multiple cameras. In Proceedings 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2003)(Cat. No. 03CH37453) (Vol. 4, pp. 3126-3131). IEEE.
  3. You, S., Tan, R. T., Kawakami, R., Mukaigawa, Y., & Ikeuchi, K. (2014, November). Raindrop detection and removal from long range trajectories. In Asian Conference on Computer Vision (pp. 569-585). Springer, Cham.
  4. You, S., Tan, R. T., Kawakami, R., Mukaigawa, Y., & Ikeuchi, K. (2015). Adherent raindrop modeling, detection and removal in video. IEEE transactions on pattern analysis and machine intelligence, 38(9), 1721-1733.
  5. Ahmed, S. T., & Patil, K. K. (2016, October). An investigative study on motifs extracted features on real time big-data signals. In 2016 International Conference on Emerging Technological Trends (ICETT) (pp. 1-4). IEEE.
  6. Halimeh, J. C., & Roser, M. (2009, June). Raindrop detection on car windshields using geometric-photometric environment construction and intensity-based correlation. In 2009 IEEE Intelligent Vehicles Symposium (pp. 610-615). IEEE.
  7. Roser, M., Kurz, J., & Geiger, A. (2010, November). Realistic modeling of water droplets for monocular adherent raindrop recognition using bezier curves. In Asian conference on computer vision (pp. 235-244). Springer, Berlin, Heidelberg.
  8. Lin, H., Jing, C., Huang, Y., & Ding, X. (2020). A 2 Net: Adjacent aggregation networks for image raindrop removal. IEEE Access, 8, 60769-60779.
  9. Periasamy, K., Periasamy, S., Velayutham, S., Zhang, Z., Ahmed, S. T., & Jayapalan, A. (2022). A proactive model to predict osteoporosis: An artificial immune system approach. Expert Systems, 39(4), e12708.
  10. Qian, R., Tan, R. T., Yang, W., Su, J., & Liu, J. (2018). Attentive generative adversarial network for raindrop removal from a single image. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 2482-2491).
  11. Al Machot, F., Ali, M., Haj Mosa, A., Schwarzlmüller, C., Gutmann, M., & Kyamakya, K. (2019). Real-time raindrop detection based on cellular neural networks for ADAS. Journal of Real-Time Image Processing, 16(4), 931-943.
  12. Alletto, S., Carlin, C., Rigazio, L., Ishii, Y., & Tsukizawa, S. (2019). Adherent raindrop removal with self-supervised attention maps and spatio-temporal generative adversarial networks. In Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops (pp. 0-0).
  13. Al-Shammari, N. K., Syed, T. H., & Syed, M. B. (2021). An Edge–IoT framework and prototype based on blockchain for smart healthcare applications. Engineering, Technology & Applied Science Research, 11(4), 7326-7331.
  14. Sreedhar Kumar, S., Ahmed, S. T., & NishaBhai, V. B. Type of Supervised Text Classification System for Unstructured Text Comments using Probability Theory Technique. International Journal of Recent Technology and Engineering (IJRTE), 8(10).
  15. Han, J., Yang, Y., Zhang, D., Huang, D., Xu, D., & De La Torre, F. (2019). Weakly-supervised learning of category-specific 3D object shapes. IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(4), 1423-1437.
  16. Zhang, D., Han, J., Yang, L., & Xu, D. (2018). SPFTN: A joint learning framework for localizing and segmenting objects in weakly labeled videos. IEEE transactions on pattern analysis and machine intelligence, 42(2), 475-489.