Contribution Analysis of Optimization Methods on Super-Resolution

Yıl 2021, Cilt: 21 Sayı: 6, 1343 - 1352, 31.12.2021

Yıldıray Anagün , Şahin Işık

https://doi.org/10.35414/akufemubid.819319

Öz

In this study, the benefits of choosing a robust optimization function with super resolution are analyzed. For this purpose, the different optimizers are included in the simple Convolutional Neural Network (CNN) architecture SRNET, to reveal the performance of the each method. Findings of this research provides that Adam and Nadam optimizers are robust when compared to (Stochastic Gradient Descent) SGD, Adagrad, Adamax and RMSprop. After experimental simulations, we have achieved the 35.91 (dB)/0.9960 and 35.97 (dB)/0.9961 accuracy rates on Set5 images from Adam and Nadam optimizers, respectively (9-1-5 network structure and filter sizes 128 and 64). These results show that selected optimization function for the CNN model plays an important role in increasing the accuracy rate in the super-resolution problem.

Anahtar Kelimeler

Super-resolution, Deep learning, Convolutional neural network, Optimizers

Optimizasyon Yöntemlerinin Süper Çözünürlük Üzerine Katkı Analizi

Öz

Bu çalışmada, süper çözünürlükte sağlam bir optimizasyon fonksiyonu seçmenin yararları analiz edilmiştir. Bu amaçla her yöntemin performansını ortaya çıkarmak için farklı optimize ediciler, basit Evrişimsel Sinir Ağı (CNN) mimarisi SRNET' e dahil edilmiştir. Bu araştırmanın bulguları, Adam ve Nadam optimize edicilerin Stokastik Gradyan İnişi (SGD), Adagrad, Adamax ve RMSprop ile karşılaştırıldığında daha kararlı olduğunu göstermektedir. Deneysel simülasyonlardan sonra, Adam ve Nadam optimize edicilerinden Set5 görüntülerinde sırasıyla 35.91 (dB)/0.9960 ve 35.97 (dB)/0.9961 doğruluk oranlarına ulaştık (9-1-5 ağ yapısı ve filtre boyutları 128 ve 64). Bu sonuçlar, CNN modeli için seçilen optimizasyon fonksiyonunun süper çözünürlük probleminde doğruluk oranını arttırmada önemli bir rol oynadığını göstermektedir.

Anahtar Kelimeler

Süper-çözünürlük, Derin öğrenme, Evrişimsel Sinir Ağı, Optimize edici

Kaynakça

• Anagun, Y., Isik, S. and Seke, E., (2019). SRLibrary: Comparing different loss functions for super-resolution over various convolutional architectures. Journal of Visual Communication and Image Representation, 61, 178-187.
• Anagün, Y., (2018). Düşük çözünürlüklü video sahnelerinden yüksek çözünürlüklü video sahnelerinin elde edilmesi. Doktora Tezi, Fen Bilimleri Enstitüsü Elektrik ve Elektronik Mühendisliği Anabilim Dalı, Eskişehir, 75.
• Bevilacqua, M., Cunningham, A., Guillemot, C., et al., (2012). Low-complexity single-image super-resolution based on nonnegative neighbor embedding.
• Dong, C., Loy, C. C., He, K., et al., (2015). Image super-resolution using deep convolutional networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 38 (2), 295-307.
• Dong, C., Loy, C. C. and Tang, X. (2016). Accelerating the Super-Resolution Convolutional Neural Network, European Conference on Computer Vision (ECCV).
• Dong, W., Zhang, L., Lukac, R., et al., (2013). Sparse representation based image interpolation with nonlocal autoregressive modeling. IEEE Trans. Image Process, 22 (4), 1382-1394.
• Dozat, T. (2016). Incorporating nesterov momentum into adam, ICLR Workshop, (1):2013–2016.
• Duchi, J., Hazan, E. and Singer, Y., (2011). Adaptive subgradient methods for online learning and stochastic optimization. 12 (7).
• Gao, X., Zhang, K. and Tao, D., (2012). Image Super-Resolution With Sparse Neighbor Embedding. IEEE Trans. Image Process, 21 (7), 3194-3205.
• He, K., Zhang, X., Ren, S., et al. (2016). Deep residual learning for image recognition, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.
• Kaveh, A. and Ezzatollah, S., (2017). Single-image super resolution using evolutionary sparse coding technique’. IET Image Process., 11 (1), 13-21.
• Kim, J., Lee, J. K. and Lee, K. M. (2016). Accurate image super-resolution using very deep convolutional networks, in IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA.
• Kingma, D. P. and Ba, J., (2014). Adam: A method for stochastic optimization.
• Lai, W. S., Huang, J. B., Ahuja, N., et al., (2018). Fast and Accurate Image Super-Resolution with Deep Laplacian Pyramid Networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 41 (11), 2599-2613.
• Ledig, C., Theis, L., Husz´ar, F., et al. (2017). Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
• Li, F., Bai, H. and Zhao, Y., (2020). Learning a Deep Dual Attention Network for Video Super-Resolution. IEEE transactions on image processing, 29, 4474-4488.
• Lim, B., Son, S., Kim, H., et al. (2017). Enhanced Deep Residual Networks for Single Image Super-Resolution, 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Honolulu, HI. Mandal, S., Bhavsar, A. and Sao, A. K., (2017). Noise adaptive super-resolution from single image via non-local mean and sparse representation. Signal Processing, 132, 134-149.
• Marivani, I., Tsiligianni, E., Cornelis, B., et al., (2020). Multimodal Deep Unfolding for Guided Image Super-Resolution. in IEEE Transactions on Image Processing, 29, 8443-8456.
• Ren, C., He, X. and Nguyen, T. Q., (2017). Single image super-resolution via adaptive high dimensional non-local total variation and adaptive geometric feature. IEEE Trans. Image Process., 26 (1), 90-106.
• Rumelhart, D. E., Hinton, G. E. and Williams, R. J., (1986). Learning representations by back-propagating errors. Nature, 323 (6088), 533-536.
• Simonyan, K. and Zisserman, A. (2015). Very deep convolutional networks for large-scale image recognition, International Conference on Learning Representations 2015 (ICLR 2015).
• Tai, Y., Yang, J. and Liu, X. (2017). Image super-resolution via deep recursive residual network, IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA.
• Tiantong, G., Hojjat, S. M. and Vishal, M., (2019). Adaptive Transform Domain Image Super-Resolution via Orthogonally Regularized Deep Networks. IEEE transactions on image processing, 28 (9), 4685-4700.
• Tieleman, T. and Hinton, G., (2012). Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude. COURSERA: Neural networks for machine learning, 4 (2), 26-31.
• Tsai, R. Y. and Huang, T. S., (1984). Multiframe image restoration and registration. Advances in Computer Vision and Image Processing. JAI Press Inc., 1 (1), 317-339.
• Yan, Q., Xu, Y., Yang, X., et al., (2015). Single image superresolution based on gradient profile sharpness. IEEE Trans. Image Process., 24 (10), 3187-3202.
• Yang, J., Wright, J., Huang, T., et al. (2008). Image super-resolution as sparse representation of raw image patches, IEEE Conference on Computer Vision and Pattern Recognition, Anchorage, AK, USA.
• Yang, J., Wright, J., Huang, T. S., et al., (2010). Image super-resolution via sparse representation. IEEE transactions on image processing, 19 (11), 2861-2873.
• Yang, M. C. and Wang, Y. C. F., (2013). A self-learning approach to single image super-reso- lution. IEEE Trans. Multimed., 15 (3), 498-508.
• Zeyde R., Elad M., and Protter M., (2010). On single image scale-up using sparse-representations. In Proceedings of the International Conference on Curves and Surfaces.
• Zhu, S., Zeng, B., Zeng, L., et al., (2016). Image interpolation based on non-local geometric similarities and directional gradients. IEEE Trans. Multimed., 18 (9), 1707-1719.