Classifying AI-Generated and Original Images Using a Convolutional Neural Network Algorithm
Abstract
This The bread industry has seen advancements in distribution, making bread products easier to transport. With the implementation of delivery orders and effective picking processes, the bread industry can streamline the shipping process and make it easier for customers to obtain bread products. To address inefficiencies in distribution, companies can manage data related to bread, orders, and sales. In this context, Toko Roti XYZ, which has three branches (Branch A, Branch B, and Branch C), faces challenges in the distribution process. The unpredictability of bread demand at these branches is a major cause of repetitive distribution. Forecasting bread demand from these branches was performed using fuzzy time series, moving average, and exponential smoothing methods to determine the quantity of products to be distributed. Subsequently, distribution optimization was carried out using the Vehicle Routing Problem (VRP) method to achieve an optimal delivery schedule that affects distribution costs with a minimum result. The forecasting results with the smallest Mean Absolute Percentage Error (MAPE) were obtained using the moving average method, with an error of 23% for large bread and 9% for small bread from Branch A, an error of 17% for large bread and 14% for small bread from Branch B, and an error of 22% for large bread and 14% for small bread from Branch C. In the VRP method, number 1 represents the depot/production place, number 2 represents Branch A, number 3 represents Branch B, and number 4 represents Branch C. The scheduling routes obtained were 1-3-2-1 and 1-4-1, with a maximum of two distributions for the two available vehicles. With these optimizations, Toko Roti XYZ was able to save distribution costs by Rp. 35,492 and reduce distribution time by 116 minutes per day compared to the previous condition. Additionally, the optimization allowed Toko Roti XYZ to reduce total carbon emissions by 5,342 kg CO₂ per year.
References
[2] Kevin Roose Sept, B. (n.d.). “I won, and I didn’t break any rules,” the artwork’s creator says. https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html
[3] Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., Liu, T., Wang, X., Wang, G., Cai, J., & Chen, T. Recent advances in convolutional neural networks. Pattern Recognition, 77, (2018) 354–377. https://doi.org/10.1016/j.patcog.2017.10.013
[4] LeCun, Y., Bengio, Y., & Hinton, G. Deep learning. Nature, 521(7553), (2015) 436–444. https://doi.org/10.1038/nature14539
[5] Bayat, O., Aljawarneh, S., Carlak, H. F., International Association of Researchers, Institute of Electrical and Electronics Engineers, & Akdeniz Üniversitesi. (2017). Proceedings of 2017 International Conference on Engineering & Technology (ICET’2017) : Akdeniz University, Antalya, Turkey, 21-23 August, 2017.
[6] Maher Salman, F., & Abu-Naser, S. S. Classification of Real and Fake Human Faces Using Deep Learning. In International Journal of Academic Engineering Research Vol. 6. (2022) www.ijeais.org/ijaer
[7] Alshariah, N. M., Khader, A., & Saudagar, J. Detecting Fake Images on Social Media using Machine Learning. In IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 10, Issue 12. (2019) www.ijacsa.thesai.org
[8] Bird, J. J., & Lotfi, A. CIFAKE: Image Classification and Explainable Identification of AI-Generated Synthetic Images. (2023) http://arxiv.org/abs/2303.14126
[9] Krizhevsky, A. Learning Multiple Layers of Features from Tiny Image. (2009)
[10] Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. High-Resolution Image Synthesis with Latent Diffusion Models. (2021) http://arxiv.org/abs/2112.10752
[11] Mascarenhas, S., & Agarwal, M. A comparison between VGG16, VGG19 and ResNet50 architecture frameworks for Image Classification. International Conference on Disruptive Technologies for Multi-Disciplinary Research and Applications (CENTCON), 1, 2021,96–99. https://doi.org/10.1109/CENTCON52345.2021.9687944
[12] Prince, S. J. D., Understanding Deep Learning (2023) https://mitpress.mit.edu,
[13] Tato, A., & Nkambou, R., Improving Adam Optimizer(2018). http://yann.lecun.com/exdb/mnist/
[14] You, Y., Zhang, Z., Hsieh, C.-J., Demmel, J., & Keutzer, K., Imagenet Training in minutes (2017). http://arxiv.org/abs/1709.05011
Authors
This work is licensed under a Creative Commons Attribution 4.0 International License.
