A Distance-Constrained Semantic Autoencoder for Zero-Shot Remote Sensing Scene Classification

Abstract

Zero-shot remote sensing scene classification refers to the classification of new images from unseen scene classes and has become a topic of growing interest in the field of remote sensing. Semantic autoencoders are one of the mainstream zeroshot learning methods. However, such autoencoders may not be discriminative enough for remote sensing scene images due to the high within-class diversity and between-class similarity. To address this issue, we propose a distance-constrained semantic autoencoder (DSAE) to deal with zero-shot remote sensing scene classification. More specifically, we learn a semantic autoencoder for seen scene classes, which allows to align the visual space and the semantic space. In order to improve the discriminative ability of this semantic autoencoder, we impose a discriminative distance metric constraint, aiming to minimize the Euclidean distances between the encoded vectors of samples of the same class and maximize the Euclidean distances between the encoded vectors of samples of different classes. Additionally, we learn a semantic autoencoder for unseen scene classes, aiming to alleviate the domain shift problem. Extensive experiments on three benchmark remote sensing scene datasets demonstrate the superiority of the proposed method over state-of-the-art methods.

Text guided zero-shot scene classification of high spatial resolution remote sensing images

“…5. Distance-constrained semantic autoencoder (DSAE) 35 adopts a distance-constrained semantic autoencoder to deal with zero-shot remote sensing scene classification.…”

“…For fair comparison, we follow the Ref. 35 to set the number of unseen classes for the zero-shot classification task.…”

“…To ensure the fairness of the comparison, the experimental settings of all methods were set using the settings in Ref. 35. To test the applicability of the proposed method for the number of unseen classes, we conduct comparison experiments under four class splits.…”

“…In Ref. 35, Wang et al. explored using the semantic autoencoder for the zero-shot remote sensing scene classification task and designed a distance-constrained semantic autoencoder.…”

Self-supervised embedding for generalized zero-shot learning in remote sensing scene classification

“…This method alters the features in the semantic space with the respective features in the visual space for maintaining the class structure consistency between visual and semantic space. Another work 24 proposes a semantic auto-encoder-based method to impose conditions on the distance to align the visual and semantic spaces for ZSL in remote sensing images. Further, a technique 25 is used to map semantic space from visual space by training a projection network to perform ZSL tasks in remote sensing images.…”