Premium
DeepSSN: A deep convolutional neural network to assess spatial scene similarity
Author(s) -
Guo Danhuai,
Ge Shiyin,
Zhang Shu,
Gao Song,
Tao Ran,
Wang Yangang
Publication year - 2022
Publication title -
transactions in gis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.721
H-Index - 63
eISSN - 1467-9671
pISSN - 1361-1682
DOI - 10.1111/tgis.12915
Subject(s) - computer science , artificial intelligence , convolutional neural network , similarity (geometry) , spatial analysis , deep learning , sketch , metric (unit) , pattern recognition (psychology) , artificial neural network , conflation , data mining , geography , image (mathematics) , remote sensing , operations management , algorithm , economics , philosophy , epistemology
Spatial‐query‐by‐sketch is an intuitive tool to explore human spatial knowledge about geographic environments and to support communication with scene database queries. However, traditional sketch‐based spatial search methods perform inadequately due to their inability to find hidden multiscale map features from mental sketches. In this research, we propose a deep convolutional neural network, namely the Deep Spatial Scene Network (DeepSSN), to better assess the spatial scene similarity. In DeepSSN, a triplet loss function is designed as a comprehensive distance metric to support the similarity assessment. A positive and negative example mining strategy is designed to ensure a consistently increasing distinction of triplets during the training process. Moreover, we develop a prototype spatial scene search system using the proposed DeepSSN, in which the users input spatial queries via sketch maps and the system can automatically augment the sketch training data. The proposed model is validated using multisource conflated map data including 131,300 labeled scene samples after data augmentation. The empirical results demonstrate that the DeepSSN outperforms baseline methods including k ‐nearest neighbors, the multilayer perceptron, AlexNet, DenseNet, and ResNet using mean reciprocal rank and precision metrics. This research advances geographic information retrieval studies by introducing a novel deep learning method tailored to spatial scene queries.