A Deep Multimodal Approach for Map Image Classification
Abstract: A large number of map images are available on the Internet, creating a growing need for efficient management, retrieval, and recommendation of such data. In this study, we train a classification model that automatically categorizes map images by theme, leveraging both textual features extracted via OCR and visual features from the images. We also release the labeled dataset constructed for our experiments to facilitate further research in this area.
Authors: Tomoya Sawada, Marie Katsurai
Publication venue: ICASSP2020
Labeled dataset
Reference
T. Sawada and M. Katsurai, “A Deep Multimodal Approach for Map Image Classification,” in 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), pp. 4457–4461, 2020. doi: 10.1109/ICASSP40776.2020.9054767. PDF