Ordinal regression

About: Ordinal regression is a research topic. Over the lifetime, 1879 publications have been published within this topic receiving 65431 citations.

Height Estimation From Single Aerial Images Using a Deep Ordinal Regression Network

Abstract: Understanding the 3-D geometric structure of the Earth's surface has been an active research topic in photogrammetry and remote sensing community for decades, serving as an essential building block for various applications such as 3-D digital city modeling, change detection, and city management. Previous research studies have extensively studied the problem of height estimation from aerial images based on stereo or multiview image matching. These methods require two or more images from different perspectives to reconstruct 3-D coordinates with camera information provided. In this letter, we deal with the ambiguous and unsolved problem of height estimation from a single aerial image. Driven by the great success of deep learning, especially deep convolutional neural networks (CNNs), some research studies have proposed to estimate height information from a single aerial image by training a deep CNN model with large-scale annotated data sets. These methods treat height estimation as a regression problem and directly use an encoder-decoder network to regress the height values. In this letter, we propose to divide height values into spacing-increasing intervals and transform the regression problem into an ordinal regression problem, using an ordinal loss for network training. To enable multiscale feature extraction, we further incorporate an Atrous Spatial Pyramid Pooling (ASPP) module to extract features from multiple dilated convolution layers. After that, a postprocessing technique is designed to transform the predicted height map of each patch into a seamless height map. Finally, we conduct extensive experiments on International Society for Photogrammetry and Remote Sensing (ISPRS) Vaihingen and Potsdam data sets. Experimental results demonstrate significantly better performance of our method compared to state-of-the-art methods.

Knee osteoarthritis severity classification with ordinal regression module

Abstract: Osteoarthritis (OA) is a common form of knee arthritis which causes significant disability and is threatening to plague patient’s quality of life. Although this chronic condition does not lead to fatality, still there exists no known cure for OA. Diagnosis of OA can be confirmed primarily based on radiographic findings. Being a progressive disease, early identification of OA is crucial for clinical interventions to curtail the OA degeneration. Kellgren-Lawrence (KL) grading system has been traditionally employed to assess the knee OA severity. Due to the recent advancements of deep learning in computer vision, more studies have employed deep neural network in automatically predicting KL grade from plain knee joint radiograph. However, these studies treat KL grading as a multi-class classification task and ignore the inherent ordinal nature within the KL grades. In this study, we propose an ordinal regression module for neural networks to treat KL grading as an ordinal regression task. Our module takes an input from neural network and produces 4 cut-points to partition the prediction space into 5 respective KL grades. The proposed model is optimized by a cumulative-link loss function. Performance of the model is evaluated against various notable neural networks and significant improvements on the knee OA KL grade prediction were demonstrated.

Joint Estimation of Age and Expression by Combining Scattering and Convolutional Networks

Abstract: This article tackles the problem of joint estimation of human age and facial expression. This is an important yet challenging problem because expressions can alter face appearances in a similar manner to human aging. Different from previous approaches that deal with the two tasks independently, our approach trains a convolutional neural network (CNN) model that unifies ordinal regression and multi-class classification in a single framework. We demonstrate experimentally that our method performs more favorably against state-of-the-art approaches.

A fast algorithm for learning large scale preference relations

Abstract: We consider the problem of learning the ranking function that maximizes a generalization of the Wilcoxon-Mann-Whitney statistic on training data. Relying on an 2-exact approximation for the error-function, we reduce the computational complexity of each iteration of a conjugate gradient algorithm for learning ranking functions from O(m2), to O(m), where m is the size of the training data. Experiments on public benchmarks for ordinal regression and collaborative filtering show that the proposed algorithm is as accurate as the best available methods in terms of ranking accuracy, when trained on the same data, and is several orders of magnitude faster.