Best Shot Selection using Convolutional Neural Networks

##plugins.themes.academic_pro.article.sidebar##

Published Jul 26, 2022
https://doi.org/10.47164/ijngc.v13i2.576
Download
Requires Subscription PDF
Statistic

Downloads

Download data is not yet available.

Metrics

Metrics Loading ...
Volume 13, Issue 2, July 2022

##plugins.themes.academic_pro.article.main##

Sagar Rane
Anant Kaulage

Abstract

Mobile phone cameras have small camera sensor size as compared to professional cameras, capture less light at a time and capture everything in focus. Also, it is very common for novice photographers to miss the best shot. These days’ companies use multiple camera systems to solve these problems, which in turn increase the complexity and the costs of deploying these systems. This paper presents an android based mobile phone camera application that takes help from the work done in segmentation networks to capture portrait images and it also tries to capture the best possible photograph using traditional image processing and convolutional neural networks. Our system is performing better in terms of Mean IOU than existing available systems. Results showed that our model is simple and cost effective. This model is easily embeddable in upcoming mobile phones as a best-shot selection feature.

##plugins.themes.academic_pro.article.details##

Creative Commons License

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

How to Cite
Rane, S., & Kaulage, A. (2022). Best Shot Selection using Convolutional Neural Networks. International Journal of Next-Generation Computing, 13(2). https://doi.org/10.47164/ijngc.v13i2.576
Crossref
Scopus
Google Scholar
Europe PMC

References

  1. abhiswain, Joe, H. 2019. Aisegment.com - matting human datasets. Kaggle.
  2. Android, S. 2018. Android. https://developer.android.com/studio.
  3. compositing, A. 2020. Alpha compositing wiki. https://en.wikipedia.org/wiki/Alphacompositing.
  4. Dlib, D. 2019. Dlib c++ library. dlib C++ Library, http://dlib.net/ .
  5. Du, X., Wang, X., Li, D., Zhu, J., Tasci, S., Upright, C., Walsh, S., and Davis, L. 2019.
  6. Boundary-sensitive network for portrait segmentation. In 2019 14th IEEE International
  7. Conference on Automatic Face Gesture Recognition (FG 2019). 1–8.
  8. Girshick, R., Donahue, J., Darrell, T., and Malik, J. 2014. Rich feature hierarchies
  9. for accurate object detection and semantic segmentation. In 2014 IEEE Conference on
  10. Computer Vision and Pattern Recognition. 580–587.
  11. Google, D. 2021. Android.hardware.camera2 android developers. Android Developers,
  12. https://developer.android.com/reference/android/hardware/camera2/package-summary.
  13. Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., An-
  14. dreetto, M., and Adam, H. 2017. Mobilenets: Efficient convolutional neural networks
  15. for mobile vision applications.
  16. Krizhevsky, A., Sutskever, I., and Hinton, G. E. 2012. Imagenet classification with deep
  17. convolutional neural networks. In Advances in Neural Information Processing Systems,
  18. F. Pereira, C. Burges, L. Bottou, and K. Weinberger, Eds. Vol. 25. Curran Associates, Inc.
  19. Labelbox. 2021. The leading training data platform for data labeling. Labelbox,
  20. https://labelbox.com/ .
  21. Mehta, S., Rastegari, M., Caspi, A., Shapiro, L., and Hajishirzi, H. 2018. Espnet:
  22. Efficient spatial pyramid of dilated convolutions for semantic segmentation.
  23. Mehta, S., Rastegari, M., Shapiro, L., and Hajishirzi, H. 2018. Espnetv2: A light-weight,
  24. power efficient, and general purpose convolutional neural network.
  25. Newbury, R., Cosgun, A., Koseoglu, M., and Drummond, T. 2019. Learning to take good
  26. pictures of people with a robot photographer.
  27. Open, Source, C. V. L. 2019. opencv. https://opencv.org/ .
  28. Park, H., Sj ̈osund, L. L., Yoo, Y., Bang, J., and Kwak, N. 2019. Extremec3net: Extreme
  29. lightweight portrait segmentation networks using advanced c3-modules.
  30. Park, H., Sj ̈osund, L. L., Yoo, Y., Monet, N., Bang, J., and Kwak, N. 2020. Sinet:
  31. Extreme lightweight portrait segmentation networks with spatial squeeze modules and in-
  32. formation blocking decoder. In 2020 IEEE Winter Conference on Applications of Computer
  33. Vision (WACV). 2055–2063.
  34. Python, Software, F. 2018. Python. https://www.python.org/ .
  35. PyTorch, L. 2020. Pytorch. https://pytorch.org/ .
  36. Redi, M., Rasiwasia, N., Aggarwal, G., and Jaimes, A. 2015. The beauty of capturing
  37. faces: Rating the quality of digital portraits.
  38. Smoothstep. 2020. Smoothstep wiki. https://en.wikipedia.org/wiki/Smoothstep.
  39. Soukupova, T. and Cech, J. 2016. Real-time eye blink detection using facial landmarks. 21st
  40. Computer Vision Winter Workshop, Slovenia.
  41. Zhang, L. and Hong, W. A. 2018. Top shot on pixel 3. Google AI Blog.
  42. Zhang, S.-H., Dong, X., Li, H., Li, R., and Yang, Y.-L. 2019. Portraitnet: Real-time
  43. portrait segmentation network for mobile device. Computers Graphics 80, 104–113.
  44. International Journal of Next-Generation Computing, Vol. 6, No. 3, November 2015.