Feature encoding models for geographic image retrieval and categorization

In this work, we survey the performance of various feature encoding models for geographic image retrieval task. Recently introduced Vector-of-Locally-Aggregated Descriptors (VLAD) and its Product Quantization encoded binary version VLAD-PQ are compared with the widely used Bag-of-Word (BoW) model. Evaluation results are shown on a publicly available 21-class LULC dataset. With experiments, it is shown that VLAD outperforms classical BoW representation albeit with some increases in the computation time. Additionally, VLAD-PQ results in similar retrieval performance with VLAD but requiring no more computational or memory resources are observed.