Divergence Regulated Encoder Network for Joint Dimensionality Reduction and Classification
Authors:
- Joshua PeeplesJPJoshua Peeples
- Sarah WalkerSWSarah Walker
- Connor McCurleyCMConnor McCurley
- Alina ZareAZAlina Zare
- James KellerJKJames Keller
- Weihuang XuWXWeihuang Xu
Preprint
Abstract
Feature representation is an important aspect of remote-sensing based image classification. While deep convolutional neural networks are able to effectively amalgamate information, large numbers of parameters often make learned features inscrutable and difficult to transfer to alternative models. In order to better represent statistical texture information for remote-sensing image classification, in this paper, we investigate performing joint dimensionality reduction and classification using a novel histogram neural network. Motivated by a popular dimensionality reduction approach, t-Distributed Stochastic Neighbor Embedding (t-SNE), our proposed method incorporates a classification loss computed on samples in a low-dimensional embedding space. We compare the learned sample embeddings against coordinates found by t-SNE in terms of classification accuracy and qualitative assessment. We also explore use of various divergence measures in the t-SNE objective. The proposed method has several advantages such as readily embedding out-of-sample points and reducing feature dimensionality while retaining class discriminability. Our results show that the proposed approach maintains and/or improves classification performance and reveals characteristics of features produced by neural networks that may be helpful for other applications.Keywords
All Related Versions
This publication has 12 references indexed in Scilit:
- Histogram Layers for Texture AnalysisIEEE Transactions on Artificial Intelligence, 2021
- Dimensionality Reduction and Classification of Hyperspectral Image via Multistructure Unified Discriminative EmbeddingIEEE Transactions on Geoscience and Remote Sensing, 2021
- Local Constraint-Based Sparse Manifold Hypergraph Learning for Dimensionality Reduction of Hyperspectral ImageIEEE Transactions on Geoscience and Remote Sensing, 2020
- CoSpace: Common Subspace Learning From Hyperspectral-Multispectral CorrespondencesIEEE Transactions on Geoscience and Remote Sensing, 2019
- Learnable manifold alignment (LeMA): A semi-supervised cross-modality learning framework for land cover and land use classificationISPRS Journal of Photogrammetry and Remote Sensing, 2019
- From BoW to CNN: Two Decades of Texture Representation for Texture ClassificationInternational Journal of Computer Vision, 2018
- Blessing of dimensionality: mathematical foundations of the statistical physics of dataPhilosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2018
- Densely Connected Convolutional Networks2017
- Closed-form supervised dimensionality reduction with generalized linear models2008
- Nonlinear Dimensionality Reduction by Locally Linear EmbeddingScience, 2000