An Experimental Study for Evaluating the Performance of CNN Pre-Trained Models in Noisy Environments

Year 2024, , 355 - 369, 29.02.2024

Halit Bakır , Sefa Burhan Eker

https://doi.org/10.2339/politeknik.1162469

Cited By: 1

Abstract

This work aims at testing the efficiency of the pre-trained models in terms of classifying images in noisy environments. To this end, we proposed injecting Gaussian noise into the images in the used datasets gradually to see how the performance of that models can be affected by the proportion of the noise in the image. Afterward, three different case studies have been conducted for evaluating the performance of six different well-known pre-trained models namely MobileNet, ResNet, GoogleNet, EfficientNet, VGG19, and Xception. In the first case study, it has been proposed to train these models using a high-quality image dataset and test them using the same datasets after injecting their images with different levels of Gaussian noise. In the second case study, we proposed training the models using the created noisy image datasets in order to investigate how the training process can be affected by the noises in the environment. In the third case study, we proposed using the non-local means algorithm to denoise the images in the noisy datasets and testing the models trained using the original datasets using these de-noised image datasets. To the best of our knowledge, this is the first time that the effects of noise on well-known pre-trained CNN architectures have been comprehensively investigated with this number of considered models. The obtained results showed that while these types of models can work very well in ideal environments their performances can drop down due to the conditions of the working environment, which reflects the need for some auxiliary models that should be used as a pre-processing phase to improve the performance of these models.

Keywords

Pre-Trained Model, Gaussian Noise, Convolutional Neural Network

Önceden Eğitilmiş CNN Modellerin Gürültülü Ortamlarda Görüntü Sınıflandırması Açısından Değerlendirilmesi

Abstract

Bu çalışma, önceden eğitilmiş CNN mimarilerinin gürültülü ortamlarda görüntüleri sınıflandırmadaki etkinliğini test etmeyi amaçlamaktadır. Bu amaçla, kullanılan veri kümelerindeki görüntülere kademeli olarak Gauss gürültüsü ekleyerek, bu modellerin performanslarının, görüntülerdeki gürültü oranından nasıl etkilenebileceğini göstermeyi hedefledik. Ardından, önceden eğitilmiş altı farklı CNN mimarisinin (MobileNet, ResNet, GoogleNet, EfficientNet, VGG19 ve Xception) performanslarını değerlendirmek için üç farklı vaka çalışması yapılmıştır. İlk vaka çalışmasında, bu mimarilerin yüksek kaliteli görüntü kümesi kullanılarak eğitilmesi, ardından aynı görüntülere farklı düzeylerde Gauss gürültüsünün enjekte edilmesi ve daha sonra gürültü içeren veri kümeleri kullanılarak bu mimarilerin test edilmesi önerilmiştir. İkinci vaka çalışmasında, CNN mimarilerindeki eğitim sürecinin ortamdaki gürültülerden nasıl etkilenebileceğini araştırmak için, oluşturulan gürültülü görüntü veri setleri kullanılarak modellerin eğitilmesi önerilmiştir. Üçüncü vaka çalışmasında ise, gürültülü veri kümelerindeki görüntülerin gürültüsünü gidermek için Non-local Means algoritmasının kullanması ve orijinal veri kümesi ile eğitilmiş modelleri, gürültüden arındırılmış veri kümeleri ile test edilmesi önerilmiştir. Bildiğimiz kadarıyla bu, önceden eğitilmiş CNN modelleri üzerinde gürültünün etkilerinin bu kadar fazla model ile deneysel olarak gösterildiği ilk çalışmadır. Elde edilen sonuçlar, bu tür modellerin ideal ortamlarda çok iyi çalışabilmelerine rağmen, gerçek hayattaki uygulamalarda çalışma ortamının koşulları nedeniyle model performanslarının düşebileceğini göstermiştir ki bu da gerçek hayattaki uygulamalarda bu modellerin performanslarını artırmak için bir ön işleme aşaması olarak kullanılması gereken bazı yardımcı modellere olan ihtiyacı göstermektedir.

Keywords

Önceden Eğitilmiş Model, Gauss Gürültüsü, Evrişimli Sinir Ağı

References

• [1] X. Lin, D. Bhattacharjee, M. el Helou, and S. Susstrunk, “Fidelity Estimation Improves Noisy-Image Classification with Pretrained Networks,” IEEE Signal Process Lett, 28: 1719–1723, (2021).
• [2] A. Awad, “Denoising images corrupted with impulse, Gaussian, or a mixture of impulse and Gaussian noise,” Engineering Science and Technology, an International Journal, 22: 746–753, (2019).
• [3] R. Rajni and A. Anutam, “Image Denoising Techniques - An Overview,” Int J Comput Appl, 86: 13–17, (2014).
• [4] M. Sheykhmousa, M. Mahdianpari, H. Ghanbari, F. Mohammadimanesh, P. Ghamisi, and S. Homayouni, “Support Vector Machine Versus Random Forest for Remote Sensing Image Classification: A Meta-Analysis and Systematic Review,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13: 6308–6325, (2020).
• [5] M. Goyal, T. Knackstedt, S. Yan, and S. Hassanpour, “Artificial intelligence-based image classification methods for diagnosis of skin cancer: Challenges and opportunities,” Computers in Biology and Medicine, 127: 104065, (2020).
• [6] X. Zhou et al., “A Comprehensive Review for Breast Histopathology Image Analysis Using Classical and Deep Neural Networks,” IEEE Access, 8: 90931–90956, (2020).
• [7] J. Naranjo-Torres, M. Mora, R. Hernández-García, R. J. Barrientos, C. Fredes, and A. Valenzuela, “A review of convolutional neural network applied to fruit image processing,” Applied Sciences (Switzerland), 10: 3443, (2020).
• [8] L. Guo, “SAR image classification based on multi-feature fusion decision convolutional neural network,” IET Image Processing, 16: 1–10, (2022).
• [9] V. D. Jan Almero, E. Sybingco, and E. P. Dadios, “An Image Classifier for Underwater Fish Detection using Classification Tree-Artificial Neural Network Hybrid; An Image Classifier for Underwater Fish Detection using Classification Tree-Artificial Neural Network Hybrid,” In2020 RIVF international conference on computing and communication technologies (RIVF), 14: 1-6, (2020).
• [10] M. Malik, F. Ahsan, and S. Mohsin, “Adaptive image denoising using cuckoo algorithm,” Soft comput, 20: 925–938, (2016).
• [11] H. R. Shahdoosti and Z. Rahemi, “Edge-preserving image denoising using a deep convolutional neural network,” Signal Processing, 159: 20–32, (2019).
• [12] K. Sirinukunwattana, S. E. A. Raza, Y. W. Tsang, D. R. J. Snead, I. A. Cree, and N. M. Rajpoot, “Locality Sensitive Deep Learning for Detection and Classification of Nuclei in Routine Colon Cancer Histology Images,” IEEE Trans Med Imaging, 35: 1196–1206, (2016).
• [13] A. Rehman, S. Naz, M. I. Razzak, F. Akram, and M. Imran, “A Deep Learning-Based Framework for Automatic Brain Tumors Classification Using Transfer Learning,” Circuits Syst Signal Process, 39: 757–775, (2020).
• [14] Y. Pathak, P. K. Shukla, A. Tiwari, S. Stalin, and S. Singh, “Deep Transfer Learning Based Classification Model for COVID-19 Disease,” IRBM, 43: 87–92, (2022).
• [15] V. K. Shrivastava and M. K. Pradhan, “Rice plant disease classification using color features: a machine learning paradigm,” Journal of Plant Pathology, 103: 17–26, (2021).
• [16] K. Thenmozhi and U. Srinivasulu Reddy, “Crop pest classification based on deep convolutional neural network and transfer learning,” Comput Electron Agric, 164:104906, (2019).
• [17] J. Wang, T. Zheng, P. Lei, and X. Bai, “Ground Target Classification in Noisy SAR Images Using Convolutional Neural Networks,” IEEE J Sel Top Appl Earth Obs Remote Sens, 11: 4180–4192, (2018).
• [18] Bakir H, Yilmaz Ş. “Using Transfer Learning Technique as a Feature Extraction Phase for Diagnosis of Cataract Disease in the Eye” International Journal of Sivas University of Science and Technology, 1: 17–33, (2022).
• [19] Doğan F, Türkoğlu İ. “Derin öğrenme algoritmalarının yaprak sınıflandırma başarımlarının karşılaştırılması” Sakarya University Journal of Computer and Information Sciences, 1: 10–21, (2018).
• [20] A. Ari and D. Hanbay, “Deep learning based brain tumor classification and detection system,” Turkish Journal of Electrical Engineering and Computer Sciences, 26: 2275–2286, (2018).
• [21] H. Firat and D. Hanbay, “3B ESA Tabanlı ResNet50 Kullanılarak Hiperspektral Görüntülerin Sınıflandırılması Classification of Hyperspectral Images Using 3D CNN Based ResNet50,” in 2021 29th Signal Processing and Communications Applications Conference (SIU), Istanbul, 6–9, (2021).
• [22] T. S. Nazaré, G. B. Costa, W. A. Contato, and M. Ponti, “Deep convolutional neural networks and noisy images,” Iberoamerican Congress on Pattern Recognition, Valparaíso, 416–424, (2017).
• [23] S. Karahan, M. K. Yildirum, K. Kirtac, F. S. Rende, G. Butun, and H. K. Ekenel, “How image degradations affect deep CNN-based face recognition?,” in 2016 international conference of the biometrics special interest group (BIOSIG), IEEE, 1–5, (2016).
• [24] A. Ali-Gombe, E. Elyan, and C. Jayne, “Fish classification in context of noisy images,” in International conference on engineering applications of neural networks, Athens, 216–226, (2017).
• [25] X. Fan et al., “Effect of image noise on the classification of skin lesions using deep convolutional neural networks,” Tsinghua Sci Technol, 25: 425–434, (2019).
• [26] K. Sriwong, K. Kerdprasop, and N. Kerdprasop, “The Study of Noise Effect on CNN-Based Deep Learning from Medical Images,” Int J Mach Learn Comput, 11: 202-207, (2021).
• [27] Bakır, Halit, and Rezan Bakır. "DroidEncoder: Malware detection using auto-encoder based feature extractor and machine learning algorithms." Computers and Electrical Engineering 110: 108804, (2023).
• [28] Bakır, Halit, Ayşe Nur Çayır, and Tuğba Selcen Navruz. "A comprehensive experimental study for analyzing the effects of data augmentation techniques on voice classification." Multimedia Tools and Applications: 1-28, (2023).
• [29] DURAN, Abdulmuttalip, and Halit BAKIR. "Hiperparametreleri Ayarlanmış Makine Öğrenimi Algoritmalarını Kullanarak Android Sistemlerde Kötü Amaçlı Yazılım Tespiti." International Journal of Sivas University of Science and Technology 2: 1-19, (2023).
• [30] Özcan, Büşra, and Halit Bakır. "YAPAY ZEKA DESTEKLİ BEYİN GÖRÜNTÜLERİ ÜZERİNDE TÜMÖR TESPİTİ." In International Conference on Pioneer and Innovative Studies, 1: 297-306, (2023).
• [31] Doğan, E. R. O. L., and Halit BAKIR. "Hiperparemetreleri Ayarlanmış Makine Öğrenmesi Yöntemleri Kullanılarak Ağdaki Saldırıların Tespiti." In International Conference on Pioneer and Innovative Studies, 1: 274-286, (2023).
• [32] Demircioğlu, Ufuk, Asaf Sayil, and Halit Bakır. "Detecting Cutout Shape and Predicting Its Location in Sandwich Structures Using Free Vibration Analysis and Tuned Machine-Learning Algorithms." Arabian Journal for Science and Engineering: 1-14, (2023).
• [33] Bakır, Halit, and Kholoud Elmabruk. "Deep learning-based approach for detection of turbulence-induced distortions in free-space optical communication links." Physica Scripta, 98: 065521, (2023).
• [34] Szegedy C, Liu W, Jia Y, Sermanet P, Reed S, Anguelov D, Erhan D, Vanhoucke V, Rabinovich A. “Going deeper with convolutions” In Proceedings of the IEEE conference on computer vision and pattern recognition, IEEE, 1-9, (2015).
• [35] Howard AG, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T, Andreetto M, Adam H. “Mobilenets: Efficient convolutional neural networks for mobile vision applications”, arXiv preprint arXiv:1704.04861, (2017).
• [36] He K, Zhang X, Ren S, Sun J. “Deep residual learning for image recognition” In Proceedings of the IEEE conference on computer vision and pattern recognition, IEEE, 770-778, (2016).
• [37] F. Chollet, “Xception: Deep learning with depthwise separable convolutions,” in Proceedings of the IEEE conference on computer vision and pattern recognition, IEEE, 1251–1258, (2017).
• [38] M. Tan and Q. Le, “Efficientnet: Rethinking model scaling for convolutional neural networks,” in International conference on machine learning, PMLR, 6105–6114, (2019).
• [39] Kazemi M, Menhaj MB. “A non-local means approach for Gaussian noise removal from images using a modified weighting kernel” arXiv preprint arXiv:1612.01006, (2016).