Açıklanabilir Yapay Zekâ Araçları ve Transfer Öğrenme ile Görüntü Tabanlı Atık Sınıflandırması

Yıl 2025, Cilt: 7 Sayı: 2, 175 - 191, 31.12.2025

İkra Kışlıoğlu , Osman Güler

https://doi.org/10.59940/jismar.1746258

Öz

Sürdürülebilir çevre yönetimi ve verimli geri dönüşüm sistemleri için etkili atık sınıflandırması esastır. Bu çalışma atıkların otomatik olarak dokuz kategoriye sınıflandırmak için RealWaste veri setinden alınan görüntü verilerini kullanan derin öğrenmeye dayalı bir yaklaşım sunmaktadır. VGG16, ResNet50, DenseNet201, EfficientNetB3 ve Xception dâhil olmak üzere on beş evrişimli sinir ağı (CNN) mimarisi, transfer öğrenmesi ve ince ayar kullanılarak değerlendirildi. Sınıf dengesizliğini azaltmak için veri artırma ve sınıf ağırlıklandırma stratejileri kullanıldı. Modeller doğruluk, kesinlik, geri çağırma, F1 puanı ve karışılık matrisleri kullanılarak değerlendirildi. Ek olarak modellerin yorumlanabilirliğini artırmak için Grad-CAM görselleştirmeleri kullanıldı. DenseNet201 ve EfficientNetB2, güçlü genelleme ve yüksek sınıf bazında performans göstererek en iyi performansı gösterenler olarak ortaya çıktı. Bu araştırma mobil veya gömülü platformlara uyarlanabilen atık sınıflandırma sistemleri oluşturmada derin öğrenmenin ve açıklanabilir yapay zekanın potansiyelini vurgulamaktadır.

Anahtar Kelimeler

Atık Sınıflandırma , Evrişimli Sinir Ağları Tranfer Öğrenme , Grad-CAM , XAI

Image-Based Waste Classification With Explainable Artificial Intelligence Tools and Transfer Learning

Öz

Effective waste classification is essential for sustainable environmental management and efficient recycling systems. This study presents a deep learning-based approach using image data from the RealWaste dataset to automatically classify waste into nine categories. Fifteen convolutional neural network (CNN) architectures, including VGG16, ResNet50, DenseNet201, EfficientNetB3, and Xception, were evaluated using transfer learning and fine-tuning. Data augmentation and class weighting strategies were employed to mitigate class imbalance. The models were assessed using accuracy, precision, recall, F1-score, and confusion matrices. Additionally, Grad-CAM visualizations were utilized to enhance the interpretability of the models. DenseNet201 and EfficientNetB2 emerged as top performers, demonstrating strong generalization and high class-wise performance. This research highlights the potential of deep learning and explainable AI in building robust waste classification systems adaptable to mobile or embedded platforms

Anahtar Kelimeler

Waste Classification , Convolutional Neural Networks , Transfer Learning , Grad-CAM , XAI

Kaynakça

• [1] M. A. Kazaure and H. Kabir, “Global waste crisis: Causes and consequences,” Environmental Sustainability Journal, vol. 9, no. 2, pp. 115–124, 2021.
• [2] J. R. Jambeck, R. Geyer, C. Wilcox, T. R. Siegler, M. Perryman, A. Andrady, and K. L. Law, “Plastic waste inputs from land into the ocean,” Science, vol. 347, no. 6223, pp. 768–771, 2015.
• [3] M. Ilyas and M. Sadiq, “An overview of municipal solid waste management in emerging countries,” Journal of Cleaner Production, vol. 265, p. 121610, 2020.
• [4] S. Pahl, K. J. Wyles, and R. C. Thompson, “Can the arts help to deliver sustainable messages?,” Philosophical Transactions of the Royal Society B, vol. 371, no. 1701, p. 20160367, 2017.
• [5]M. Önder, N. Daldal, K. Polat, and M. U. Doğan, “Internet of thing-based hand tremor monitoring system and automated detection of hand tremor frequency,” Computers and Electrical Engineering, vol. 122, 2025.
• [6] Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, no. 7553, pp. 436–444, 2015.
• [7] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet classification with deep convolutional neural networks,” Communications of the ACM, vol. 60, no. 6, pp. 84–90, 2017.
• [8] S. J. Pan and Q. Yang, “A survey on transfer learning,” IEEE Transactions on Knowledge and Data Engineering, vol. 22, no. 10, pp. 1345–1359, 2010.
• [9] A. Adadi and M. Berrada, “Peeking inside the black-box: A survey on explainable artificial intelligence (XAI),” IEEE Access, vol. 6, pp. 52138–52160, 2018.
• [10] H. Al-Mashhadani, "Waste classification based on deep learning algorithms," Journal of Intelligent Systems, vol. 32, no. 1, pp. 321–332, 2023. doi: 10.1515/jisys-2023-0064
• [11] A. A. Younis and A. A. Obaid, "Intelligent waste classification using EfficientNet based deep learning," IJACSA, vol. 15, no. 11, pp. 486–492, 2024.
• [12] R. Qiu, Y. Zhou, J. Wang, and Y. Li, "Real-time waste classification using EfficientNetV2," arXiv preprint, arXiv:2503.21208, 2025.
• [13] B. Yalçın, Z. Cebeci, and R. Altıoklar, "Deep learning-based waste classification system using RealWaste dataset," Information, vol. 14, no. 12, p. 633, 2023.
• [14] A. Kunwar, N. Gupta, and S. Sharma, "Intelligent plastic waste classification using YOLO-11m," arXiv preprint, arXiv:2412.20232, 2024.
• [15] S. Narayan, "DeepWaste: A mobile app for real-time waste classification using deep learning," arXiv preprint, arXiv:2101.05960, 2021.
• [16] R. R. Selvaraju et al., "Grad-CAM: Visual explanations from deep networks via gradient-based localization," in Proc. IEEE Int. Conf. Comput. Vis., 2017, pp. 618–626.
• [17] M. Mazloumian, N. Amini, and M. Derakhshan, "Food waste detection and classification using CNNs," arXiv preprint, arXiv:2002.03786, 2020.
• [18] J. Guo, Y. Xu, and H. Wu, "Waste image classification using superpixel and deep convolutional neural networks," IJCRT, vol. 8, no. 6, pp. 1856–1862, 2020.
• [19] V. Badrinarayanan, A. Kendall, and R. Cipolla, "SegNet: A deep convolutional encoder-decoder architecture for image segmentation," IEEE TPAMI, vol. 39, no. 12, pp. 2481–2495, 2017.
• [20] A. A. Alade, O. O. Jegede, and O. A. Fakorede, "Deep learning in waste management: A brief survey," Int. J. Comput. Anal. Strategy Tech., vol. 14, no. 2, pp. 101–116, 2024.
• [21] M. Tan and Q. V. Le, "EfficientNet: Rethinking model scaling for convolutional neural networks," in Proc. ICML, 2019, pp. 6105–6114.
• [22] K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," arXiv preprint, arXiv:1409.1556, 2014.
• [23] K. He, X. Zhang, S. Ren, and J. Sun, "Deep residual learning for image recognition," in Proc. IEEE CVPR, 2016, pp. 770–778.
• [24] C. Shorten and T. M. Khoshgoftaar, "A survey on image data augmentation for deep learning," Journal of Big Data, vol. 6, no. 1, pp. 1–48, 2019.
• [25] S. Single, S. Iranmanesh, and R. Raad, "Realwaste: A novel real-life data set for landfill waste classification using deep learning," Information, vol. 14, no. 12, p. 633, 2023. doi: 10.3390/info14120633