VivahPix: Intelligent Wedding Photo Organization Using Deep Learning-Based Event Classification and Automated Album Generation

This paper presented VivahPix, an intelligent wedding photo organization framework that integrates image preprocessing, deep learning-based event classification, and automated album generation. The system employs duplicate image detection using perceptual hashing (pHash), blur detection using the Variance of Laplacian method, and CNN-based classification to automatically categorize photographs into Engagement, Haldi, Mehndi, Wedding, Reception, and Miscellaneous events.     A comparative evaluation of AlexNet, VGG19, MobileNetV2, InceptionV3, and ResNet50 demonstrated that ResNet50 achieved the highest classification accuracy of 96%. The proposed hybrid feature extraction approach also achieved superior Precision, Recall, and F1-Score values of 96%, while data augmentation further improved model robustness and generalization capability.     The framework successfully automated wedding photograph organization by generating event-wise albums and separating duplicate and blurry images, significantly reducing manual sorting effort. Deployment as a web-based application further demonstrated its practical applicability in real-world scenarios.     Future work may focus on incorporating transformer-based vision models, facial recognition for person-wise grouping, video-based event classification, automated highlight generation from wedding videos, image caption generation, and personalized album design recommendations. The framework can also be extended to process both photographs and videos, enabling comprehensive multimedia wedding album management with improved intelligence, scalability, and user experience.

[1]           P. Shivakumara, C. P. Kumar, J. J. Nemade, K. Michael, A. Kumar, B. S. Anami, and U. Pal, “A new U-Net based system for multi-cultural wedding image classification,” Expert Systems with Applications, vol. 237, p. 121562, 2024, doi: 10.1016/j.eswa.2023.121562. [2]           S. Khara, “Artificial intelligence and deep learning in wedding photo and video content creation,” Zenodo Preprint, 2024, doi: 10.5281/zenodo.15168074. [3]           M. H. Hadid, Q. M. Hussein, Z. T. Al-Qaysi, M. A. Ahmed, and M. M. Salih, “An overview of content-based image retrieval methods and techniques,” International Journal of Computer Science and Mathematics, vol. 3, no. 2, pp. 40–51, Jul. 2023, doi: 10.52866/ijcsm.2023.02.03.006. [4]           M. Tripathi, “Analysis of convolutional neural network based image classification techniques,” Journal of Innovative Image Processing, vol. 3, no. 2, pp. 100–117, Jun. 2021, doi: 10.1016/j.procs.2018.05.198. [5]           L. M. Francis and N. Sreenath, “Pre-processing techniques for detection of blurred images,” in Proc. Int. Conf. on Computational Intelligence and Data Engineering (ICCIDE), Lecture Notes on Data Engineering and Communications Technologies, vol. 28, Springer, 2019, pp. 59–66, doi: 10.1007/978-981-13-6459-4_7. [6]           S. R. Dubey, “A decade survey of content-based image retrieval using deep learning,” arXiv preprint arXiv:2012.00641v1, Nov. 2020, doi: 10.1109/TCSVT.2021.3080920. [7]           X. Li, J. Yang, and J. Ma, “Recent developments of content-based image retrieval (CBIR),” Neurocomputing, vol. 452, pp. 423–448, 2021, doi: 10.1016/j.neucom.2020.07.139. [8]           M. N. Abdullah, M. A. M. Shukran, M. R. M. Isa, N. S. M. Ahmad, M. A. Khairuddin, M. S. F. M. Yunus, and F. Ahmad, “Colour feature extraction techniques and approaches for content-based image retrieval (CBIR) systems,” Journal of Materials Science and Chemical Engineering, vol. 9, pp. 29–34, Jul. 2021, doi: 10.4236/msce.2021.97003. [9]           M. Buvana, K. Muthumayil, and T. Jayasankar, “Content-based image retrieval based on hybrid feature extraction and feature selection technique using pigeon inspired optimization,” Annals of R.S.C.B., vol. 25, no. 1, pp. 424–443, 2021. [10]         S. R. Dubey, “A decade survey of content-based image retrieval using deep learning,” arXiv preprint arXiv:2010.01313, 2020, doi: 10.1109/TCSVT.2021.3080920.