Fine-Grained Object Recognition and Zero-Shot Learning in Remote Sensing Imagery

TL;DR

This paper presents a zero-shot learning approach for fine-grained object recognition in high-resolution aerial imagery, enabling the identification of unseen classes using semantic auxiliary information and a new dataset of street trees.

Contribution

It introduces a novel zero-shot learning framework for remote sensing imagery and a new dataset with auxiliary attributes, improving recognition of unseen classes.

Findings

Achieved 14.3% accuracy on unseen classes, outperforming random guess and other ZSL methods.

Demonstrated the effectiveness of auxiliary information like attributes, language models, and taxonomy.

Provided a new dataset with 40 street tree types at 1-ft resolution for fine-grained recognition tasks.

Abstract

Fine-grained object recognition that aims to identify the type of an object among a large number of subcategories is an emerging application with the increasing resolution that exposes new details in image data. Traditional fully supervised algorithms fail to handle this problem where there is low between-class variance and high within-class variance for the classes of interest with small sample sizes. We study an even more extreme scenario named zero-shot learning (ZSL) in which no training example exists for some of the classes. ZSL aims to build a recognition model for new unseen categories by relating them to seen classes that were previously learned. We establish this relation by learning a compatibility function between image features extracted via a convolutional neural network and auxiliary information that describes the semantics of the classes of interest by using training samples from the seen classes. Then, we show how knowledge transfer can be performed for the unseen classes by maximizing this function during inference. We introduce a new data set that contains 40 different types of street trees in 1-ft spatial resolution aerial data, and evaluate the performance of this model with manually annotated attributes, a natural language model, and a scientific taxonomy as auxiliary information. The experiments show that the proposed model achieves 14.3% recognition accuracy for the classes with no training examples, which is significantly better than a random guess accuracy of 6.3% for 16 test classes, and three other ZSL algorithms.