Transformer and VOTE Based AI Model for Waste Classification: An Application of Recycling in Smart Cities

Abstract

As industrialization and smart city development activities increase, waste collection, classification, and planning have become important. The recycling process of waste is based on recovering its properties as they are in their natural locations, reducing pollution, and helping to create a sustainable environment. This study aims to improve the performance of deep learning models for converting organic waste into recyclable waste. In the study, a transformer-based hybrid model was proposed for waste classification. The proposed model used DinoV2, ConvNeXtV2, and ViT-B16 as the base models. Feature maps derived using these models were concatenated and then classified into different classifiers. In the last step of our model, the features obtained from different classifiers were subjected to the VOTE hard-voting method. The proposed model achieved 96.64% success in solid waste classification. The VOTE process improved the proposed model's performance from 95.58% to 96.64%. When the features obtained from the transformer-based models used in the study were classified by the classifiers, the highest accuracy, 95.37%, was achieved with the DinoV2 model. The values obtained in the proposed model are important for the future of smart and sustainable cities.

Keywords

• Waste Separation
• Classification
• Transformers
• Recycling
• Smart Cities

Atık Sınıflandırması için Dönüştürücü ve Oylama Tabanlı YZ Modeli: Akıllı Şehirlerde Geri Dönüşümün Bir Uygulaması

Abstract

Sanayileşme ve akıllı şehir geliştirme faaliyetleri arttıkça, atık toplama, sınıflandırma ve planlama önem kazanmıştır. Atık geri dönüşüm süreci, atıkların doğal ortamlarındaki özelliklerini geri kazanmaya, kirliliği azaltmaya ve sürdürülebilir bir çevre oluşturmaya dayanmaktadır. Bu çalışma, organik atıkları geri dönüştürülebilir atıklara dönüştürmek için derin öğrenme modellerinin performansını iyileştirmeyi amaçlamaktadır. Bunun için çalışmada RealWaste veri seti kullanılmış ve sonuçlar bu veri setinde alınmıştır. Çalışmada, atık sınıflandırması için transformatör tabanlı hibrit bir model önerilmiştir. Önerilen model, temel modeller olarak DinoV2, ConvNeXtV2 ve ViT-B16'yı kullanmıştır. Bu modeller kullanılarak elde edilen özellik haritaları birleştirilmiş ve daha sonra farklı sınıflandırıcılar halinde sınıflandırılmıştır. Modelimizin son adımında, farklı sınıflandırıcılardan elde edilen özellikler, VOTE sert oylama yöntemine tabi tutulmuştur. Önerilen model, katı atık sınıflandırmasında %96,64 başarı elde etmiştir. VOTE süreci, önerilen modelin performansını %95,58'den %96,64'e yükseltmiştir. Çalışmada kullanılan transformatör tabanlı modellerden elde edilen özellikler sınıflandırıcılar tarafından sınıflandırıldığında, en yüksek doğruluk oranı olan %95,37, DinoV2 modeli ile elde edilmiştir. Önerilen modelde elde edilen değerler, akıllı ve sürdürülebilir şehirlerin geleceği için önemlidir.

Keywords

• Atık ayrıştırma
• sınıflandırma
• transformatörler
• geri dönüşüm
• akıllı şehirler

References

1. D. Szpilko, A. de la Torre Gallegos, F. Jimenez Naharro, A. Rzepka, A. Remiszewska, Waste management in the smart city: current practices and future directions, Resources 12 (2023) 115.
2. D. Singh, A.K. Dikshit, S. Kumar, Smart technological options in collection and transportation of municipal solid waste in urban areas: A mini review, Waste Manag. \& Res. 42 (2024) 3–15.
3. A. Lakhouit, Revolutionizing Urban Solid Waste Management with AI and IoT: A review of smart solutions for waste collection, sorting, and recycling, Results Eng. (2025) 104018.
4. K. Anokye, A.S. Mohammed, P. Agyemang, A.B. Agya, E.E.Y. Amuah, S. Sodoke, A systematic review of the impacts of open burning and open dumping of waste in Ghana: A way forward for sustainable waste management, Clean. Waste Syst. (2024) 100152.
5. V. Kanna, S. Roseline, K. Balamurugan, S. Jeeva, I. Augastin Santhiyagu, The effects of greenhouse gas emissions on global warming, Encycl. Renew. Energy, Sustain. Environ. 1 (2024) 143–154.
6. P. Ubolsook, C. Podong, S. Sedpho, P. Jansanthea, Assessing the environmental impact of construction waste management in northern Thailand: An approach to estimate greenhouse gas emissions and cumulative energy demand, J. Clean. Prod. 467 (2024) 142961.
7. I. Aiguobarueghian, U.M. Adanma, E.O. Ogunbiyi, N.O. Solomon, Waste management and circular economy: A review of sustainable practices and economic benefits, World J. Adv. Res. Rev. 22 (2024) 1708–1719.
8. R.R. Karri, N.M. Mubarak, S.K. Bhagat, T. Tiyasha, L.P. Lingamdinne, J.R. Koduru, G. Ravindran, I. Tyagi, M.H. Dehghani, Scientometrics and overview of water, environment, and sustainable development goals, in: Water, Environ. Sustain. Dev. Goals, Elsevier, 2024: pp. 3–33.

9. H. Sharma, H. Kumar, A computer vision-based system for real-time component identification from waste printed circuit boards, J. Environ. Manage. 351 (2024) 119779.
10. T. Ji, J. Li, H. Fang, R. Zhang, J. Yang, L. Fan, Rapid dataset generation methods for stacked construction solid waste based on machine vision and deep learning, PLoS One 19 (2024) e0296666.
11. S. Sriram, D.K. Singh, D. V Sairam, V. Nivethitha, M. Thangavel, Anterior Cruciate Ligament (ACL) Tear Detection using Hybrid CNN Transformer, IEEE Access (2025).
12. B. Kelenyi, V. Domsa, L. Tamas, SAM-Net: self-attention based feature matching with spatial transformers and knowledge distillation, Expert Syst. Appl. 242 (2024) 122804.
13. Z. Lu, F. Wang, Z. Xu, F. Yang, T. Li, On the performance and memory footprint of distributed training: An empirical study on transformers, Softw. Pract. Exp. (2024).
14. Z. Zhao, T. Chen, J. Dou, G. Liu, A. Plaza, Landslide susceptibility mapping considering landslide local-global features based on CNN and transformer, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. (2024).
15. M.M. Hossen, M.E. Majid, S.B.A. Kashem, A. Khandakar, M. Nashbat, A. Ashraf, M. Hasan-Zia, A.K.A. Kunju, S. Kabir, M.E.H. Chowdhury, A reliable and robust deep learning model for effective recyclable waste classification, IEEE Access 12 (2024) 13809–13821.
16. H. Zhang, H. Cao, Y. Zhou, C. Gu, D. Li, Hybrid deep learning model for accurate classification of solid waste in the society, Urban Clim. 49 (2023) 101485.
17. M.I.B. Ahmed, R.B. Alotaibi, R.A. Al-Qahtani, R.S. Al-Qahtani, S.S. Al-Hetela, K.A. Al-Matar, N.K. Al-Saqer, A. Rahman, L. Saraireh, M. Youldash, others, Deep learning approach to recyclable products classification: Towards sustainable waste management, Sustainability 15 (2023) 11138.
18. S. Jin, Z. Yang, G. Królczykg, X. Liu, P. Gardoni, Z. Li, Garbage detection and classification using a new deep learning-based machine vision system as a tool for sustainable waste recycling, Waste Manag. 162 (2023) 123–130.
19. S. Single, S. Iranmanesh, R. Raad, Realwaste: A novel real-life data set for landfill waste classification using deep learning, Information 14 (2023) 633.
20. U.K. Lilhore, S. Simaiya, S. Dalal, R. Damaševičius, A smart waste classification model using hybrid CNN-LSTM with transfer learning for sustainable environment, Multimed. Tools Appl. 83 (2024) 29505–29529.
21. M.M. Hossen, A. Ashraf, M. Hasan, M.E. Majid, M. Nashbat, S.B.A. Kashem, A.K.A. Kunju, A. Khandakar, S. Mahmud, M.E.H. Chowdhury, GCDN-Net: Garbage classifier deep neural network for recyclable urban waste management, Waste Manag. 174 (2024) 439–450.
22. M. Chhabra, B. Sharan, M. Elbarachi, M. Kumar, Intelligent waste classification approach based on improved multi-layered convolutional neural network, Multimed. Tools Appl. 83 (2024) 84095–84120.
23. F.S. Alrayes, M.M. Asiri, M.S. Maashi, M.K. Nour, M. Rizwanullah, A.E. Osman, S. Drar, A.S. Zamani, Waste classification using vision transformer based on multilayer hybrid convolution neural network, Urban Clim. 49 (2023) 101483.
24. R. Pitakaso, T. Srichok, S. Khonjun, P. Golinska-Dawson, K. Sethanan, N. Nanthasamroeng, S. Gonwirat, P. Luesak, C. Boonmee, Optimization-driven artificial intelligence-enhanced municipal waste classification system for disaster waste management, Eng. Appl. Artif. Intell. 133 (2024) 108614.
25. Ogut, Z., Karaduman, M., Bozdag, P. G., Karakose, M., & Yildirim, M. (2026). A Hybrid Model with Quantum Feature Map Based on CNN and Vision Transformer for Clinical Support in Diagnosis of Acute Appendicitis. Biomedicines, 14(1), 183.
26. Ogut, Z., Karaduman, M., & Yildirim, M. (2025). Clinically Focused Computer-Aided Diagnosis for Breast Cancer Using SE and CBAM with Multi-Head Attention. Tomography, 11(12), 138.
27. Kızılkaya, A. R., Karaduman, M., Kurt, H., Yıldırım, T. T., Düger, N., Talo, F., & Yıldırım, M. (2026). Automatic Detection of Dental Implant Brands Using a Hybrid Vision Transformer-Based Deep Learning Model. Journal of Oral Implantology, 1(aop).
28. K. He, C. Gan, Z. Li, I. Rekik, Z. Yin, W. Ji, Y. Gao, Q. Wang, J. Zhang, D. Shen, Transformers in medical image analysis, Intell. Med. 3 (2023) 59–78.
29. S. Hameed, J. Amin, M.A. Anjum, M. Sharif, Suspicious activities detection using spatial--temporal features based on vision transformer and recurrent neural network, J. Ambient Intell. Humaniz. Comput. 15 (2024) 3379–3391.
30. Z. Cai, L. Lin, H. He, X. Tang, Uni4eye: Unified 2d and 3d self-supervised pre-training via masked image modeling transformer for ophthalmic image classification, in: Int. Conf. Med. Image Comput. Comput. Interv., 2022: pp. 88–98.
31. A. Shonenkov, A. Kuznetsov, D. Dimitrov, T. Shavrina, D. Chesakov, A. Maltseva, A. Fenogenova, I. Pavlov, A. Emelyanov, S. Markov, others, RuCLIP--new models and experiments: a technical report, ArXiv Prepr. ArXiv2202.10784 (2022).
32. S.S. Mahanty, D. Muduli, A. Kumari, S.K. Sharma, Pretrained DeIT for Brain Tumor Classification: A Fine-Tuning Approach with Label Smoothing, in: 2024 15th Int. Conf. Comput. Commun. Netw. Technol., 2024: pp. 1–6.
33. M. Oquab, T. Darcet, T. Moutakanni, H. Vo, M. Szafraniec, V. Khalidov, P. Fernandez, D. Haziza, F. Massa, A. El-Nouby, others, Dinov2: Learning robust visual features without supervision, ArXiv Prepr. ArXiv2304.07193 (2023).
34. M. Elhenawy, H.I. Ashqar, A. Rakotonirainy, T.I. Alhadidi, A. Jaber, M.A. Tami, Vision-Language Models for Autonomous Driving: CLIP-Based Dynamic Scene Understanding, Electronics 14 (2025) 1282.
35. B. Ozdemir, I. Pacal, An innovative Deep learning framework for skin cancer detection employing ConvNeXtV2 and focal self-attention mechanisms, Results Eng. 25 (2025) 103692.
36. G. Guo, H. Wang, D. Bell, Y. Bi, K. Greer, KNN model-based approach in classification, in: Move to Meaningful Internet Syst. 2003 CoopIS, DOA, ODBASE OTM Confed. Int. Conf. CoopIS, DOA, ODBASE 2003, Catania, Sicily, Italy, Novemb. 3-7, 2003. Proc., 2003: pp. 986–996.
37. S.V.M. Vishwanathan, M.N. Murty, SSVM: a simple SVM algorithm, in: Proc. 2002 Int. Jt. Conf. Neural Networks. IJCNN’02 (Cat. No. 02CH37290), 2002: pp. 2393–2398.
38. B. De Ville, Decision trees, Wiley Interdiscip. Rev. Comput. Stat. 5 (2013) 448–455.
39. P. Xanthopoulos, P.M. Pardalos, T.B. Trafalis, P. Xanthopoulos, P.M. Pardalos, T.B. Trafalis, Linear discriminant analysis, Robust Data Min. (2013) 27–33.
40. T.R. Patil, S.S. Sherekar, others, Performance analysis of Naive Bayes and J48 classification algorithm for data classification, Int. J. Comput. Sci. Appl. 6 (2013) 256–261.
41. A. Dogan, D. Birant, A weighted majority voting ensemble approach for classification, in: 2019 4th Int. Conf. Comput. Sci. Eng., 2019: pp. 1–6.
42. M. Al-Raeei, The smart future for sustainable development: Artificial intelligence solutions for sustainable urbanization, Sustain. Dev. 33 (2025) 508–517.