MyHistory: photo gallery management

Abstract

In today’s digital era, the ease of capturing photos has led to an overwhelming accumulation of images on personal devices, making organization and storage management increasingly challenging. People frequently take multiple shots of the same moment to capture the best photo, resulting in cluttered galleries that are difficult to sort. To address these issues, this project presents MyHistory, an intelligent photo management application designed to automate organization, enhance photo selection, and minimize unnecessary storage usage. The system enables users to create meaningful albums based on keywords and metadata, allowing for efficient photo categorization. In addition, the system provides a nearduplicate photo recommendation feature, which assess image aesthetics and applies human-centric criteria to recommend the best photos to keep. Beyond organization and recommendations, the system introduces an automated collage generation feature, enabling users to create visually appealing photo collages effortlessly. Users can select from multiple collage styles, while the system intelligently arranges high-quality images based on aesthetic scores. The prototype iOS application successfully integrates these features, demonstrating a practical and efficient approach to digital photo management. By combining metadatadriven album creation, deep learning-based classification, and advanced recommendation techniques, MyHistory provides a seamless and user-friendly experience, helping users preserve and relive their memories with ease.