Abstract:Conventional application of convolutional neural networks (CNNs) for image classification and recognition is based on the assumption that all target classes are equal(i.e., no hierarchy) and exclusive of one another (i.e., no overlap). CNN-based image classifiers built on this assumption, therefore, cannot take into account an innate hierarchy among target classes (e.g., cats and dogs in animal image classification) or additional information that can be easily derived from the data (e.g.,numbers larger than five in the recognition of handwritten digits), thereby resulting in scalability issues when the number of target classes is large. Combining two related but slightly different ideas of hierarchical classification and logical learning by auxiliary inputs, we propose a new learning framework called hierarchical auxiliary learning, which not only address the scalability issues with a large number of classes but also could further reduce the classification/recognition errors with a reasonable number of classes. In the hierarchical auxiliary learning, target classes are semantically or non-semantically grouped into superclasses, which turns the original problem of mapping between an image and its target class into a new problem of mapping between a pair of an image and its superclass and the target class. To take the advantage of superclasses, we introduce an auxiliary block into a neural network, which generates auxiliary scores used as additional information for final classification/recognition; in this paper, we add the auxiliary block between the last residual block and the fully-connected output layer of the ResNet. Experimental results demonstrate that the proposed hierarchical auxiliary learning can reduce classification errors up to 0.56, 1.6 and 3.56 percent with MNIST, SVHN and CIFAR-10 datasets, respectively.

Constructing Class Hierarchy via Adaptive Kernel Learning.

Adaptive kernel learning based on centered alignment for hierarchical classification

Building Hierarchical Class Structures for Extreme Multi-Class Learning

Efficient discriminative learning of class hierarchy for many class prediction

Exploring and Exploiting Hierarchical Structures for Large-Scale Classification

Hierarchical Multiple Kernel Clustering

Hierarchical Fine-Grained Visual Classification Leveraging Consistent Hierarchical Knowledge

Clustering with Adaptive Structure Learning: A Kernel Approach

Data-driven hierarchical structure kernel for multiscale part-based object recognition

Label Hierarchy Transition: Modeling Class Hierarchies to Enhance Deep Classifiers

Hierarchical Auxiliary Learning

Hierarchical Abstracting Graph Kernel

Fast Meta-Learning for Adaptive Hierarchical Classifier Design

Adaptive Construction of Hierarchical Classifiers Basing on the Confusion Relationship among Multiple Patterns

MMF: Multi-Task Multi-Structure Fusion for Hierarchical Image Classification

Efficient Heuristics for Structure Learning of k -Dependence Bayesian Classifier.

Harnessing Superclasses for Learning from Hierarchical Databases

Feature selection via maximizing inter-class independence and minimizing intra-class redundancy for hierarchical classification.

Hierarchical Structure Construction on Hypergraphs

Discriminative Fast Hierarchical Learning for Multiclass Image Classification

Performance Improvement in Multi-class Classification via Automated Hierarchy Generation and Exploitation through Extended LCPN Schemes