Image Retrieval for Local Architectural Heritage Recommendation Based on Deep Hashing

Propagating architectural heritage is of great significance to the inheritance and protection local culture. Recommendations based on user preferences can greatly benefit promotion so as better protect inherit historical Thus, a powerful tool necessary build such recommendation system. Recently, deep learning methods have proliferated means analyze data in domains. In this paper, case study Jiangxi, China, we explore system for area. To organize our experiments, dataset traditional Chinese architecture constructed hashing retrieval method proposed task. By utilizing fine-tuning strategy, realize high-accuracy break model training restriction caused by insufficient heritage. Furthermore, answers map into two-dimensional space reveal relationships between different categories. An image-to-location application also provided experience.