Supervised Learning Framework Research Articles

Health assessment of high-speed train wheels based on group-profile data

The rapid development of high-speed trains has brought a significant demand to increase the reliability and optimize the maintenance of train wheels. As the state-of-the-art practice in high-speed trains, the maximal radial run-out and equivalent conicity are two leading health indicators (HIs) to assess the health status of the wheels. However, these two HIs cannot effectively assess the degree of wheel polygonal wear, which has been associated with the service failure of structural components. In the article, we propose a data-driven supervised learning framework for extracting a multi-dimensional HI to assess the condition of the wheels using group-profile data. To the authors ' knowledge, it is the first proposed multi-dimensional HI for the high-speed train wheels. The proposed framework is based on the proper integration of feature extraction and regression techniques, e.g., Hilbert-Huang transform, Functional Principal Component Analysis, and Logistic Regression. A set of real-world high-speed train wheel profile data are collected to validate the proposed framework. The statistical results show that the HI generated from the proposed framework outperforms the traditional HIs in abnormal wheels detection, i.e., classification. Additionally, the conditional probability based on the wheel profile data is proposed in this paper to achieve condition-based maintenance.

Partial Label Learning with competitive learning graph neural network

Partial Label Learning (PLL) is a weakly supervised learning framework where each instance may be associated with more than one candidate label, among which only one is true. Traditionally, the PLL problem is solved by removing the false candidate labels based on the instance relationship, while the potentially useful information between instances and labels as well as the potential candidate label relationship is ignored. In this paper, a new PLL framework PL-CGNN is proposed, which treats the instances with false labels as noise, and the PLL is reformulated to remove the noise instances. First of all, the feature of each label class is approximately represented by the center point of all the related instances. The significant operation enables the similarity between instances and labels measurable. Next, all the candidate labels for each instance compete for the biggest similarity. To further improve the robustness of the model, the competition procedure for the most similar label is extended to the neighbors of this instance. The label with the most wins is the final ground-truth one. The relationship between candidate labels guides the situation that the competition process develops into. Through iterative competitive learning, each label class approaches the true value. Experiments carried out on diverse datasets show that the performance of the PL-CGNN model is outstanding.

Evidence Transfer: Learning Improved Representations According to External Heterogeneous Task Outcomes

Unsupervised representation learning tends to produce generic and reusable latent representations. However, these representations can often miss high-level features or semantic information, since they only observe the implicit properties of the dataset. On the other hand, supervised learning frameworks learn task-oriented latent representations that may not generalise in other tasks or domains. In this article, we introduce evidence transfer, a deep learning method that incorporates the outcomes of external tasks in the unsupervised learning process of an autoencoder. External task outcomes also referred to as categorical evidence, are represented by categorical variables, and are either directly or indirectly related to the primary dataset—in the most straightforward case they are the outcome of another task on the same dataset. Evidence transfer allows the manipulation of generic latent representations in order to include domain or task-specific knowledge that will aid their effectiveness in downstream tasks. Evidence transfer is robust against evidence of low quality and effective when introduced with related, corresponding, or meaningful evidence.

Partial Label Dimensional Reduction via Semantic Difference Information and Manifold Regularization

Partial label learning is a rising weakly supervised learning framework that deals with the problem that each training instance is associated with a set of candidate labels, where only one is correct. The dimensionality reduction mechanism can effectively improve the learning system’s generalization performance. However, due to the ambiguity of the ground-truth label, it is hard to use traditional dimensionality reduction methods directly on partially labeled data sets. Existing research attempts to exploit the linear discriminant analysis (LDA) strategy to reduce the dimensionality of partial label training examples. Unfortunately, this method still suffered from false-positive labels. This paper proposed an LDA-based dimensionality reduction approach named SDIMR for partially labeled datasets. On the one hand, introduce the manifold regularization term with semantic difference information to perform dimensionality reduction procedures and simultaneously maintain the local manifold structure. On the other hand, the iterative process alternates between dimensionality reduction and disambiguation. Briefly, our approach consists of three stages: Firstly, utilize the graph structure with semantic difference information to describe the topological information in the feature space. Secondly, we propose a new objective function containing manifold regularization terms to reduce the dimensionality of original training data to generate a new training data set. Finally, the k-nearest neighbor method is employed to disambiguate the dimensionality-reduced data. Extensive experiments on artificial and real-world data sets show that our method can be comparable to state-of-the-art partial label learning algorithms.

Weakly-supervised learning for catheter segmentation in 3D frustum ultrasound.

Accurate and efficient catheter segmentation in 3D ultrasound (US) is essential for ultrasound-guided cardiac interventions. State-of-the-art segmentation algorithms, based on convolutional neural networks (CNNs), suffer from high computational cost and large 3D data size for GPU implementation, which are far from satisfactory for real-time applications. In this paper, we propose a novel approach for efficient catheter segmentation in 3D US. Instead of using Cartesian US, our approach performs catheter segmentation in Frustum US (i.e., the US data before scan conversion). Compared to Cartesian US, Frustum US has a much smaller volume size, therefore the catheter can be segmented more efficiently in Frustum US. However, annotating the irregular and deformed Frustum images is challenging, and it is laborious to obtain the voxel-level annotation. To address this, we propose a weakly supervised learning framework, which requires only bounding-box annotations. The labels of the voxels are generated by incorporating class activation maps with line filtering, which are iteratively updated during the training cycles. Our experimental results show that, compared to Cartesian US, the catheter can be segmented much more efficiently in Frustum US (i.e., 0.25s per volume) with better accuracy. Extensive experiments also validate the effectiveness of the proposed weakly supervised learning method.

SIFLoc: a self-supervised pre-training method for enhancing the recognition of protein subcellular localization in immunofluorescence microscopic images.

With the rapid growth of high-resolution microscopy imaging data, revealing the subcellular map of human proteins has become a central task in the spatial proteome. The cell atlas of the Human Protein Atlas (HPA) provides precious resources for recognizing subcellular localization patterns at the cell level, and the large-scale annotated data enable learning via advanced deep neural networks. However, the existing predictors still suffer from the imbalanced class distribution and the lack of labeled data for minor classes. Thus, it is necessary to develop new methods for coping with these issues. We leverage the self-supervised learning protocol to address these problems. Especially, we propose a pre-training scheme to enhance the conventional supervised learning framework called SIFLoc. The pre-training is featured by a hybrid data augmentation method and a modified contrastive loss function, aiming to learn good feature representations from microscopic images. The experiments are performed on a large-scale immunofluorescence microscopic image dataset collected from the HPA database. Using the same deep neural networks as the classifier, the model pre-trained via SIFLoc not only outperforms the model without pre-training by a large margin but also shows advantages over the state-of-the-art self-supervised learning methods. Especially, SIFLoc improves the prediction accuracy for minor organelles significantly.

Regression and Classification With Spline-Based Separable Expansions.

We introduce a supervised learning framework for target functions that are well approximated by a sum of (few) separable terms. The framework proposes to approximate each component function by a B-spline, resulting in an approximant where the underlying coefficient tensor of the tensor product expansion has a low-rank polyadic decomposition parametrization. By exploiting the multilinear structure, as well as the sparsity pattern of the compactly supported B-spline basis terms, we demonstrate how such an approximant is well-suited for regression and classification tasks by using the Gauss–Newton algorithm to train the parameters. Various numerical examples are provided analyzing the effectiveness of the approach.

Deep convolutional neural network for multi-level non-invasive tunnel lining assessment

In recent years, great attention has focused on the development of automated procedures for infrastructures control. Many efforts have aimed at greater speed and reliability compared to traditional methods of assessing structural conditions. The paper proposes a multi-level strategy, designed and implemented on the basis of periodic structural monitoring oriented to a cost- and time-efficient tunnel control plan. Such strategy leverages the high capacity of convolutional neural networks to identify and classify potential critical situations. In a supervised learning framework, Ground Penetrating Radar (GPR) profiles and the revealed structural phenomena have been used as input and output to train and test such networks. Image-based analysis and integrative investigations involving video-endoscopy, core drilling, jacking and pull-out testing have been exploited to define the structural conditions linked to GPR profiles and to create the database. The degree of detail and accuracy achieved in identifying a structural condition is high. As a result, this strategy appears of value to infrastructure managers who need to reduce the amount and invasiveness of testing, and thus also to reduce the time and costs associated with inspections made by highly specialized technicians.

A GNN-Based Supervised Learning Framework for Resource Allocation in Wireless IoT Networks

The Internet of Things (IoT) allows physical devices to be connected over the wireless networks. Although device-to-device (D2D) communication has emerged as a promising technology for IoT, the conventional solutions for D2D resource allocation are usually computationally complex and time consuming. The high complexity poses a significant challenge to the practical implementation of wireless IoT networks. A graph neural network (GNN)-based framework is proposed to address this challenge in a supervised manner. Specifically, the wireless network is modeled as a directed graph, where the desirable communication links are modeled as nodes and the harmful interference links are modeled as edges. The effectiveness of the proposed framework is verified via two case studies, namely the link scheduling in D2D networks and the joint channel and power allocation in D2D underlaid cellular networks. Simulation results demonstrate that the proposed framework outperforms the benchmark schemes in terms of the average sum rate and the sample efficiency. In addition, the proposed GNN approach shows potential generalizability to different system settings and robustness to the corrupted input features. It also accelerates the D2D resource optimization by reducing the execution time to only a few milliseconds.

Similar Pair-free Partial Label Metric Learning

Partial label learning is a new weak- ly supervised learning framework. In this frame- work, the real category label of a training sample is usually concealed in a set of candidate labels, which will lead to lower accuracy of learning al- gorithms compared with traditional strong super- vised cases. Recently, it has been found that met- ric learning technology can be used to improve the accuracy of partial label learning algorithm- s. However, because it is difficult to ascertain similar pairs from training samples, at present there are few metric learning algorithms for par- tial label learning framework. In view of this, this paper proposes a similar pair-free partial la- bel metric learning algorithm. The main idea of the algorithm is to define two probability distri- butions on the training samples, i.e., the proba- bility distribution determined by the distance of sample pairs and the probability distribution de- termined by the similarity of candidate label set of sample pairs, and then the metric matrix is ob- tained via minimizing the KL divergence of the two probability distributions. The experimental results on several real-world partial label dataset- s show that the proposed algorithm can improve the accuracy of k-nearest neighbor partial label learning algorithm (PL-KNN) better than the ex- isting partial label metric learning algorithms, up to 8 percentage points.

Disambiguation Enabled Linear Discriminant Analysis for Partial Label Dimensionality Reduction

As an emerging weakly supervised learning framework, partial label learning considers inaccurate supervision where each training example is associated with multiple candidate labels among which only one is valid. In this article, a first attempt toward employing dimensionality reduction to help improve the generalization performance of partial label learning system is investigated. Specifically, the popular linear discriminant analysis (LDA) techniques are endowed with the ability of dealing with partial label training examples. To tackle the challenge of unknown ground-truth labeling information, a novel learning approach named Delin is proposed which alternates between LDA dimensionality reduction and candidate label disambiguation based on estimated labeling confidences over candidate labels. On one hand, the (kernelized) projection matrix of LDA is optimized by utilizing disambiguation-guided labeling confidences. On the other hand, the labeling confidences are disambiguated by resorting to k NN aggregation in the LDA-induced feature space. Extensive experiments over a broad range of partial label datasets clearly validate the effectiveness of Delin in improving the generalization performance of well-established partial label learning algorithms.

Integrating Character-level and Word-level Representation for Affect in Arabic Tweets

Affect tasks, which range from sentiment polarity classification to finer grained sentiment strength and emotional intensity detection, have become of increasing interest due to the vast amount of user-generated content and advanced learning models. Word representation models have been leveraged effectively within a variety of natural language processing tasks. However, these models are not always effective in the context of social media. When dealing with social media posts in Arabic, the use of Arabic dialects needs to be considered. Although using informal text to train word-level models can lead to the identification of words that convey the same meaning, these models are unable to capture the full extent of the words that are used in the real world due to out-of-vocabulary (OOV) words. The inability to identify such words is one of the main limitations of word-level models. One approach of overcoming OOV is through the use of character-level embeddings as they can effectively learn the vectors of word parts or character n-grams. This study uses a combination of character-level and word-level models to identify the most effective methods by which affective Arabic words in tweets can be represented semantically and morphologically. We evaluate our generated models and the proposed method by integrating them in a supervised learning framework that was used for a range of affect tasks and other related tasks. Our findings reveal that the developed models surpassed the performance of state-of-the-art Arabic pre-trained word embeddings over eight datasets. In addition, our models enhance previous state-of-the-art outcomes on tasks involving Arabic emotion intensity, outperforming the top-systems that used advanced ensemble learning models and several additional features.

Weakly Supervised Spatial Deep Learning for Earth Image Segmentation Based on Imperfect Polyline Labels

In recent years, deep learning has achieved tremendous success in image segmentation for computer vision applications. The performance of these models heavily relies on the availability of large-scale high-quality training labels (e.g., PASCAL VOC 2012). Unfortunately, such large-scale high-quality training data are often unavailable in many real-world spatial or spatiotemporal problems in earth science and remote sensing (e.g., mapping the nationwide river streams for water resource management). Although extensive efforts have been made to reduce the reliance on labeled data (e.g., semi-supervised or unsupervised learning, few-shot learning), the complex nature of geographic data such as spatial heterogeneity still requires sufficient training labels when transferring a pre-trained model from one region to another. On the other hand, it is often much easier to collect lower-quality training labels with imperfect alignment with earth imagery pixels (e.g., through interpreting coarse imagery by non-expert volunteers). However, directly training a deep neural network on imperfect labels with geometric annotation errors could significantly impact model performance. Existing research that overcomes imperfect training labels either focuses on errors in label class semantics or characterizes label location errors at the pixel level. These methods do not fully incorporate the geometric properties of label location errors in the vector representation. To fill the gap, this article proposes a weakly supervised learning framework to simultaneously update deep learning model parameters and infer hidden true vector label locations. Specifically, we model label location errors in the vector representation to partially reserve geometric properties (e.g., spatial contiguity within line segments). Evaluations on real-world datasets in the National Hydrography Dataset (NHD) refinement application illustrate that the proposed framework outperforms baseline methods in classification accuracy.

Unsupervised Cycle-Consistent Generative Adversarial Networks for Pan Sharpening

Deep learning based pan-sharpening has received significant research interest in recent years. Most of existing methods fall into the supervised learning framework in which they down-sample the multi-spectral (MS) and panchromatic (PAN) images and regard the original MS images as ground truths to form training samples. Although impressive performance could be achieved, they have difficulties generalizing to the original full-scale images due to the scale gap, which makes them lack of practicability. In this paper, we propose an unsupervised generative adversarial framework that learns from the full-scale images without the ground truths to alleviate this problem. We extract the modality-specific features from the PAN and MS images with a two-stream generator, perform fusion in the feature domain, and then reconstruct the pan-sharpened images. Furthermore, we introduce a novel hybrid loss based on the cycle-consistency and adversarial scheme to improve the performance. Comparison experiments with the state-of-the-art methods are conducted on GaoFen-2 and WorldView-3 satellites. Results demonstrate that the proposed method can greatly improve the pan-sharpening performance on the full-scale images, which clearly show its practical value. Codes are available at https://github.com/zhysora/UCGAN.

3D Layout Estimation via Weakly Supervised Learning of Plane Parameters From 2D Segmentation.

The task of 3D layout estimation in an indoor scene is to predict the holistic 3D structural information of the scene from an RGB image. It is costly to obtain the ground truth 3D layout, and this issue severely restricts the learning based 3D layout estimation approaches. In this paper, we present a novel weakly supervised learning framework that is able to learn the 3D layout effectively with 2D layout segmentation mask as supervision. We employ a deep neural network to predict the plane parameters and camera intrinsic parameters in the image. Based on the predicted plane instances, the 3D layout as well as the corresponding depth map and 2D segmentation can be generated. The key objectives for learning meaningful plane parameters are the label consistency of layout segmentation and depth consistency of border pixels from adjacent planes, with which the ground truth 2D layout segmentation is able to supervise the learning of the 3D layout. We further incorporate 3D geometric reasoning and prior knowledge in the learning process to ensure that the learned 3D layout is realistic and reasonable. Experimental results show that our method can produce accurate 3D layout estimates by weakly supervised learning.

Learning Pain from Action Unit Combinations: A Weakly Supervised Approach via Multiple Instance Learning.

Patient pain can be detected highly reliably from facial expressions using a set of facial muscle-based action units (AUs) defined by the Facial Action Coding System (FACS). A key characteristic of facial expression of pain is the simultaneous occurrence of pain-related AU combinations, whose automated detection would be highly beneficial for efficient and practical pain monitoring. Existing general Automated Facial Expression Recognition (AFER) systems prove inadequate when applied specifically for detecting pain as they either focus on detecting individual pain-related AUs but not on combinations or they seek to bypass AU detection by training a binary pain classifier directly on pain intensity data but are limited by lack of enough labeled data for satisfactory training. In this paper, we propose a new approach that mimics the strategy of human coders of decoupling pain detection into two consecutive tasks: one performed at the individual video-frame level and the other at video-sequence level. Using state-of-the-art AFER tools to detect single AUs at the frame level, we propose two novel data structures to encode AU combinations from single AU scores. Two weakly supervised learning frameworks namely multiple instance learning (MIL) and multiple clustered instance learning (MCIL) are employed corresponding to each data structure to learn pain from video sequences. Experimental results show an 87% pain recognition accuracy with 0.94 AUC (Area Under Curve) on the UNBC-McMaster Shoulder Pain Expression dataset. Tests on long videos in a lung cancer patient video dataset demonstrates the potential value of the proposed system for pain monitoring in clinical settings.

Deep Bilateral Filtering Network for Point-Supervised Semantic Segmentation in Remote Sensing Images.

Semantic segmentation methods based on deep neural networks have achieved great success in recent years. However, training such deep neural networks relies heavily on a large number of images with accurate pixel-level labels, which requires a huge amount of human effort, especially for large-scale remote sensing images. In this paper, we propose a point-based weakly supervised learning framework called the deep bilateral filtering network (DBFNet) for the semantic segmentation of remote sensing images. Compared with pixel-level labels, point annotations are usually sparse and cannot reveal the complete structure of the objects; they also lack boundary information, thus resulting in incomplete prediction within the object and the loss of object boundaries. To address these problems, we incorporate the bilateral filtering technique into deeply learned representations in two respects. First, since a target object contains smooth regions that always belong to the same category, we perform deep bilateral filtering (DBF) to filter the deep features by a nonlinear combination of nearby feature values, which encourages the nearby and similar features to become closer, thus achieving a consistent prediction in the smooth region. In addition, the DBF can distinguish the boundary by enlarging the distance between the features on different sides of the edge, thus preserving the boundary information well. Experimental results on two widely used datasets, the ISPRS 2-D semantic labeling Potsdam and Vaihingen datasets, demonstrate that our proposed DBFNet can achieve a highly competitive performance compared with state-of-the-art fully-supervised methods. Code is available at https://github.com/Luffy03/DBFNet.

An Ensemble Supervised Learning Framework Based on Latent Representation Learning for Incomplete and High-Dimensional Data

An Ensemble Supervised Learning Framework Based on Latent Representation Learning for Incomplete and High-Dimensional Data

Fast Successive-Cancellation List Flip Decoding of Polar Codes

This work presents a fast successive-cancellation list flip (Fast-SCLF) decoding algorithm for polar codes that addresses the high latency issue associated with the successive-cancellation list flip (SCLF) decoding algorithm. We first propose a bit-flipping strategy tailored to the state-of-the-art fast successive-cancellation list (FSCL) decoding that avoids tree-traversal in the binary tree representation of SCLF, thus reducing the latency of the decoding process. We then derive a parameterized path selection error model to accurately estimate the bit index at which the correct decoding path is eliminated from the initial FSCL decoding. The trainable parameter is optimized online based on an efficient supervised learning framework. Simulation results show that for a polar code of length 512 with 256 information bits, with similar error-correction performance and memory consumption, the proposed Fast-SCLF decoder reduces up to $73.4\%$ of the average decoding latency of the SCLF decoder with the same list size at the frame error rate of $10^{-4}$, while incurring a maximum computational complexity overhead of $27.6\%$. For the same polar code of length 512 with 256 information bits and at practical signal-to-noise ratios, the proposed decoder with list size 4 reduces $89.3\%$ and $43.7\%$ of the average complexity and decoding latency of the FSCL decoder with list size 32 (FSCL-32), respectively, while also reducing $83.2\%$ of the memory consumption of FSCL-32. The significant improvements of the proposed decoder come at the cost of $0.07$ dB error-correction performance degradation compared with FSCL-32.

A Supervised Learning Framework for Joint Estimation of Angles-of-Arrival and Number of Sources

Machine learning is a promising technique for angle-of-arrival (AOA) estimation of waves impinging a sensor array. However, the majority of the methods proposed so far only consider a known, fixed number of impinging waves, i.e., a fixed number of sources (NOS). This paper proposes a machine-learning-based estimator designed for the case when the NOS is variable and hence unknown a priori. The proposed estimator comprises a framework of single-label classifiers. Each classifier predicts if waves are present within certain randomly selected segments of the array’s field of view (FOV), resulting from discretising the FOV with a certain (FOV) resolution. The classifiers’ predictions are combined into a probabilistic angle spectrum, whereupon the NOS and the AOAs are estimated jointly by applying a probability threshold whose optimal level is learned from data. The estimator’s performance is assessed using a new performance metric: the joint AOA estimation success rate. Numerical simulations show that for low SNR (−10 dB), a low FOV resolution (2°) yields a higher success rate than a high resolution (1°), whereas the opposite applies for mid (0 dB) and high (10 dB) SNRs. In nearly all simulations, except one at low SNR and a high FOV resolution, the proposed estimator outperforms the MUSIC algorithm if the maximum allowed AOA estimation error is approximately equal to (or larger than) the FOV resolution.