Data-efficient image classification, which focuses on achieving accurate classification performance with limited labeled data, has garnered significant attention. Genetic programming (GP) has achieved impressive progress in image classification, particularly in scenarios involving small amounts of labeled data. GP research typically focuses on designing tree-based model representations to learn useful image features for classification. However, most GP methods are proposed for gray-scale images and ignore the color features. Furthermore, the existing GP methods typically learn features on a single scale/resolution, restricting potential accuracy enhancements. To address these issues, this paper proposes a new multi-tree GP In single-tree GP (or simply GP), each individual consists of a single tree. In contrast, in multi-tree GP, each individual comprises multiple trees. representation for image feature learning and classification. In each individual, three trees are included to extract discriminative features from the red, green, and blue channels of the image. With the new image resizing layer in the tree representation, the proposed approach can achieve multi-scale feature extraction, i.e., flexibly learning fine-grained details and coarse-grained structures in the image, improving the classification performance. In addition, since a limitation of GP is premature convergence due to a decline in population diversity, this paper develops a hybrid parent selection method consisting of tournament and lexicase selection to increase population diversity, find the best individual, and improve classification accuracy. The experiments on six image classification datasets indicate that the proposed approach outperforms state-of-the-art neural network-based and GP-based methods in almost all comparisons. Further analyses demonstrate the effectiveness of each component and the potentially high interpretability of the proposed approach.

中文翻译：

---

用于学习图像分类中的颜色和多尺度特征的多树遗传编程


数据高效的图像分类专注于利用有限的标记数据实现准确的分类性能，已经引起了人们的广泛关注。遗传编程（GP）在图像分类方面取得了令人瞩目的进展，特别是在涉及少量标记数据的场景中。 GP 研究通常侧重于设计基于树的模型表示，以学习有用的图像特征以进行分类。然而，大多数GP方法都是针对灰度图像提出的，而忽略了颜色特征。此外，现有的 GP 方法通常在单一尺度/分辨率上学习特征，限制了潜在的准确性增强。为了解决这些问题，本文提出了一种新的多树GP。在单树GP（或简称GP）中，每个个体都由一棵树组成。相反，在多树GP中，每个个体都包含多棵树。图像特征学习和分类的表示。每个个体中都包含三棵树，用于从图像的红色、绿色和蓝色通道中提取判别特征。通过树表示中新的图像调整层，该方法可以实现多尺度特征提取，即灵活地学习图像中的细粒度细节和粗粒度结构，提高分类性能。此外，由于GP的局限性是由于种群多样性下降而导致过早收敛，因此本文开发了一种由锦标赛和词汇选择组成的混合父选择方法，以增加种群多样性，找到最佳个体，提高分类精度。 对六个图像分类数据集的实验表明，所提出的方法在几乎所有比较中都优于最先进的基于神经网络和基于 GP 的方法。进一步的分析证明了每个组件的有效性以及所提出方法的潜在高可解释性。