Semi-supervised Deep Learning Research Articles

Accurate recognition of colorectal cancer with semi-supervised deep learning on pathological images.

Machine-assisted pathological recognition has been focused on supervised learning (SL) that suffers from a significant annotation bottleneck. We propose a semi-supervised learning (SSL) method based on the mean teacher architecture using 13,111 whole slide images of colorectal cancer from 8803 subjects from 13 independent centers. SSL (~3150 labeled, ~40,950 unlabeled; ~6300 labeled, ~37,800 unlabeled patches) performs significantly better than the SL. No significant difference is found between SSL (~6300 labeled, ~37,800 unlabeled) and SL (~44,100 labeled) at patch-level diagnoses (area under the curve (AUC): 0.980 ± 0.014 vs. 0.987 ± 0.008, P value = 0.134) and patient-level diagnoses (AUC: 0.974 ± 0.013 vs. 0.980 ± 0.010, P value = 0.117), which is close to human pathologists (average AUC: 0.969). The evaluation on 15,000 lung and 294,912 lymph node images also confirm SSL can achieve similar performance as that of SL with massive annotations. SSL dramatically reduces the annotations, which has great potential to effectively build expert-level pathological artificial intelligence platforms in practice.

Deep semi-supervised learning ensemble framework for classifying co-mentions of human proteins and phenotypes

BackgroundIdentifying human protein-phenotype relationships has attracted researchers in bioinformatics and biomedical natural language processing due to its importance in uncovering rare and complex diseases. Since experimental validation of protein-phenotype associations is prohibitive, automated tools capable of accurately extracting these associations from the biomedical text are in high demand. However, while the manual annotation of protein-phenotype co-mentions required for training such models is highly resource-consuming, extracting millions of unlabeled co-mentions is straightforward.ResultsIn this study, we propose a novel deep semi-supervised ensemble framework that combines deep neural networks, semi-supervised, and ensemble learning for classifying human protein-phenotype co-mentions with the help of unlabeled data. This framework allows the ability to incorporate an extensive collection of unlabeled sentence-level co-mentions of human proteins and phenotypes with a small labeled dataset to enhance overall performance. We develop PPPredSS, a prototype of our proposed semi-supervised framework that combines sophisticated language models, convolutional networks, and recurrent networks. Our experimental results demonstrate that the proposed approach provides a new state-of-the-art performance in classifying human protein-phenotype co-mentions by outperforming other supervised and semi-supervised counterparts. Furthermore, we highlight the utility of PPPredSS in powering a curation assistant system through case studies involving a group of biologists.ConclusionsThis article presents a novel approach for human protein-phenotype co-mention classification based on deep, semi-supervised, and ensemble learning. The insights and findings from this work have implications for biomedical researchers, biocurators, and the text mining community working on biomedical relationship extraction.

A semi-supervised deep learning model for ship encounter situation classification

Maritime safety is an important issue for global shipping industries. Currently, most of collision accidents at sea are caused by the misjudgement of the ship’s operators. The deployment of maritime autonomous surface ships (MASS) can greatly reduce ships’ reliance on human operators by using an automated intelligent collision avoidance system to replace human decision-making. To successfully develop such a system, the capability of autonomously identifying other ships and evaluating their associated encountering situation is of paramount importance. In this paper, we aim to identify ships’ encounter situation modes using deep learning methods based upon the Automatic Identification System (AIS) data. First, a segmentation process is developed to divide each ship’s AIS data into different segments that contain only one encounter situation mode. This is different to the majority of studies that have proposed encounter situation mode classification using hand-crafted features, which may not reflect the actual ship’s movement states. Furthermore, a number of present classification tasks are conducted using substantial labelled AIS data followed by a supervised training paradigm, which is not applicable to our dataset as it contains a large number of unlabelled AIS data. Therefore, a method called Semi-Supervised Convolutional Encoder–Decoder Network (SCEDN) for ship encounter situation classification based on AIS data is proposed. The structure of the network is not only able to automatically extract features from AIS segments but also share training parameters for the unlabelled data. The SCEDN uses an encoder–decoder convolutional structure with four channels for each segment (distance, speed, Time to the Closed Point of Approach (TCPA) and Distance to the Closed Point of Approach (DCPA)) been developed. The performance of the SCEDN model are evaluated by comparing to several baselines with the experimental results demonstrating a higher accuracy can be achieved by our proposed model.

Deep semi-supervised learning of dynamics for anomaly detection in laser powder bed fusion

Highly complex data streams from in-situ additive manufacturing (AM) monitoring systems are becoming increasingly prevalent, yet finding physically actionable patterns remains a key challenge. Recent AM literature utilising machine learning methods tend to make predictions about flaws or porosity without considering the dynamical nature of the process. This leads to increases in false detections as useful information about the signal is lost. This study takes a different approach and investigates learning a physical model of the laser powder bed fusion process dynamics. In addition, deep representation learning enables this to be achieved directly from high speed videos. This representation is combined with a predictive state space model which is learned in a semi-supervised manner, requiring only the optimal laser parameter to be characterised. The model, referred to as FlawNet, was exploited to measure offsets between predicted and observed states resulting in a highly robust metric, known as the dynamic signature. This feature also correlated strongly with a global material quality metric, namely porosity. The model achieved state-of-the-art results with a receiver operating characteristic (ROC) area under curve (AUC) of 0.999 when differentiating between optimal and unstable laser parameters. Furthermore, there was a demonstrated potential to detect changes in ultra-dense, 0.1% porosity, materials with an ROC AUC of 0.944, suggesting an ability to detect anomalous events prior to the onset of significant material degradation. The method has merit for the purposes of detecting out of process distributions, while maintaining data efficiency. Subsequently, the generality of the methodology would suggest the solution is applicable to different laser processing systems and can potentially be adapted to a number of different sensing modalities.

Automatic Classification of Ovarian Cancer Types from CT Images Using Deep Semi-Supervised Generative Learning and Convolutional Neural Network

Automatic Classification of Ovarian Cancer Types from CT Images Using Deep Semi-Supervised Generative Learning and Convolutional Neural Network

Anomaly Detection on Wind Turbines Based on a Deep Learning Analysis of Vibration Signals

ABSTRACT In this paper, we present a Semi-Supervised Deep Learning approach for anomaly detection of Wind Turbine generators based on vibration signals. The proposed solution is integrated into an IoT Platform as a real-time Workflow. The Workflow is responsible for the whole detection process when a new sample is inserted in the IoT Platform, performing data gathering, preprocessing, feature extraction, and classification. The chosen Semi-Supervised Deep Learning model is a DAE, which builds a normality model using healthy data. The classification consists of comparing the reconstruction error for the computed entry with a normality threshold. The normality threshold is selected through an F1-Score analysis of the reconstruction errors over labeled data. Finally, the Workflow can produce notifications to the users whenever unhealthy behavior is noticed. The ability of the proposed mechanism to detect abnormal behavior in wind turbines on an IoT Platform is evaluated using a case study of real-world healthy and unhealthy data from a Wind Turbine. The solution was able to correctly classify every unhealthy sample and presented a low false-positive rate. Moreover, Workflow results can be improved by conditioning alarm triggering with a windowed-based anomaly accumulation.

Semi-HIC: A novel semi-supervised deep learning method for histopathological image classification

Histopathological images provide a gold standard for cancer recognition and diagnosis. Existing approaches for histopathological image classification are supervised learning methods that demand a large amount of labeled data to obtain satisfying performance, which have to face the challenge of limited data annotation due to prohibitive time cost. To circumvent this shortage, a promising strategy is to design semi-supervised learning methods. Recently, a novel semi-supervised approach called Learning by Association (LA) is proposed, which achieves promising performance in nature image classification. However, there are still great challenges in its application to histopathological image classification due to the wide inter-class similarity and intra-class heterogeneity in histopathological images. To address these issues, we propose a novel semi-supervised deep learning method called Semi-HIC for histopathological image classification. Particularly, we introduce a new semi-supervised loss function combining an association cycle consistency (ACC) loss and a maximal conditional association (MCA) loss, which can take advantage of a large number of unlabeled patches and address the problems of inter-class similarity and intra-class variation in histopathological images, and thereby remarkably improve classification performance for histopathological images. Besides, we employ an efficient network architecture with cascaded Inception blocks (CIBs) to learn rich and discriminative embeddings from patches. Experimental results on both the Bioimaging 2015 challenge dataset and the BACH dataset demonstrate our Semi-HIC method compares favorably with existing deep learning methods for histopathological image classification and consistently outperforms the semi-supervised LA method.

Semi-supervised target classification in multi-frequency echosounder data

Abstract Acoustic target classification in multi-frequency echosounder data is a major interest for the marine ecosystem and fishery management since it can potentially estimate the abundance or biomass of the species. A key problem of current methods is the heavy dependence on the manual categorization of data samples. As a solution, we propose a novel semi-supervised deep learning method leveraging a few annotated data samples together with vast amounts of unannotated data samples, all in a single model. Specifically, two inter-connected objectives, namely, a clustering objective and a classification objective, optimize one shared convolutional neural network in an alternating manner. The clustering objective exploits the underlying structure of all data, both annotated and unannotated; the classification objective enforces a certain consistency to given classes using the few annotated data samples. We evaluate our classification method using echosounder data from the sandeel case study in the North Sea. In the semi-supervised setting with only a tenth of the training data annotated, our method achieves 67.6% accuracy, outperforming a conventional semi-supervised method by 7.0 percentage points. When applying the proposed method in a fully supervised setup, we achieve 74.7% accuracy, surpassing the standard supervised deep learning method by 4.7 percentage points.

Adaptable and Explainable Predictive Maintenance: Semi-Supervised Deep Learning for Anomaly Detection and Diagnosis in Press Machine Data

Predictive maintenance (PdM) has the potential to reduce industrial costs by anticipating failures and extending the work life of components. Nowadays, factories are monitoring their assets and most collected data belong to correct working conditions. Thereby, semi-supervised data-driven models are relevant to enable PdM application by learning from assets’ data. However, their main challenges for application in industry are achieving high accuracy on anomaly detection, diagnosis of novel failures, and adaptability to changing environmental and operational conditions (EOC). This article aims to tackle these challenges, experimenting with algorithms in press machine data of a production line. Initially, state-of-the-art and classic data-driven anomaly detection model performance is compared, including 2D autoencoder, null-space, principal component analysis (PCA), one-class support vector machines (OC-SVM), and extreme learning machine (ELM) algorithms. Then, diagnosis tools are developed supported on autoencoder’s latent space feature vector, including clustering and projection algorithms to cluster data of synthetic failure types semi-supervised. In addition, explainable artificial intelligence techniques have enabled to track the autoencoder’s loss with input data to detect anomalous signals. Finally, transfer learning is applied to adapt autoencoders to changing EOC data of the same process. The data-driven techniques used in this work can be adapted to address other industrial use cases, helping stakeholders gain trust and thus promote the adoption of data-driven PdM systems in smart factories.

Ensemble-based approach for semisupervised learning in remote sensing

Semisupervised learning (SSL) techniques explore the progressive discovery of the hidden latent data structure by propagating supervised information on unlabeled data, which are thereafter used to reinforce learning. These schemes are beneficial in remote sensing, where thousands of new images are added every day, and manual labeling results are prohibitive. Our work introduces an ensemble-based semisupervised deep learning approach that initially takes a subset of labeled data Dl, which represents the latent structure of the data and progressively propagates labels automatically from an expanding set of unlabeled data Du. The ensemble is a set of classifiers whose predictions are collated to derive a consolidated prediction. Only those data having a high-confidence prediction are considered as newly generated labels. The proposed approach was exhaustively validated on four public datasets, achieving appreciable results compared to the state-of-the-art methods in most of the evaluated configurations. For all datasets, the proposed approach achieved a classification F1-score and recall of up to 90%, on average. The SSL and recursive scheme also demonstrated an average gain of ∼2 % at the last training stage in such large datasets.

Semi-Supervised Spatiotemporal Deep Learning for Intrusions Detection in IoT Networks

The rapid growth of the Internet of Things (IoT) technologies has generated a huge amount of traffic that can be exploited for detecting intrusions through IoT networks. Despite the great effort made in annotating IoT traffic records, the number of labeled records is still very small, increasing the difficulty in recognizing attacks and intrusions. This study introduces a semi-supervised deep learning approach for intrusion detection (SS-Deep-ID), in which we propose a multiscale residual temporal convolutional (MS-Res) module to finetune the network capability in learning spatiotemporal representations. An improved traffic attention (TA) mechanism is introduced to estimate the importance score that helps the model to concentrate on important information during learning. Furthermore, a hierarchical semi-supervised training method is introduced which takes into account the sequential characteristics of the IoT traffic data during training. The proposed SS-Deep-ID is easily integrated into a fog-enabled IoT network to offer efficient real-time intrusion detection. Finally, empirical evaluations on two recent data sets (CIC-IDS2017 and CIC-IDS2018) demonstrate that SS-Deep-ID improves the efficiency of intrusion detection and increases the robustness of performance while maintaining computational efficiency.

Paddy Rice Imagery Dataset for Panicle Segmentation

Accurate panicle identification is a key step in rice-field phenotyping. Deep learning methods based on high-spatial-resolution images provide a high-throughput and accurate solution of panicle segmentation. Panicle segmentation tasks require costly annotations to train an accurate and robust deep learning model. However, few public datasets are available for rice-panicle phenotyping. We present a semi-supervised deep learning model training process, which greatly assists the annotation and refinement of training datasets. The model learns the panicle features with limited annotations and localizes more positive samples in the datasets, without further interaction. After the dataset refinement, the number of annotations increased by 40.6%. In addition, we trained and tested modern deep learning models to show how the dataset is beneficial to both detection and segmentation tasks. Results of our comparison experiments can inspire others in dataset preparation and model selection.

Semi-Supervised Deep Learning for Lunar Crater Detection Using CE-2 DOM

Lunar craters are very important for estimating the geological age of the Moon, studying the evolution of the Moon, and for landing site selection. Due to a lack of labeled samples, processing times due to high-resolution imagery, the small number of suitable detection models, and the influence of solar illumination, Crater Detection Algorithms (CDAs) based on Digital Orthophoto Maps (DOMs) have not yet been well-developed. In this paper, a large number of training data are labeled manually in the Highland and Maria regions, using the Chang’E-2 (CE-2) DOM; however, the labeled data cannot cover all kinds of crater types. To solve the problem of small crater detection, a new crater detection model (Crater R-CNN) is proposed, which can effectively extract the spatial and semantic information of craters from DOM data. As incomplete labeled samples are not conducive for model training, the Two-Teachers Self-training with Noise (TTSN) method is used to train the Crater R-CNN model, thus constructing a new model—called Crater R-CNN with TTSN—which can achieve state-of-the-art performance. To evaluate the accuracy of the model, three other detection models (Mask R-CNN, no-Mask R-CNN, and Crater R-CNN) based on semi-supervised deep learning were used to detect craters in the Highland and Maria regions. The results indicate that Crater R-CNN with TTSN achieved the highest precision (of 91.4% and 88.5%, respectively) in the Highland and Maria regions, even obtaining the highest recall and F1 score. Compared with Mask R-CNN, no-Mask R-CNN, and Crater R-CNN, Crater R-CNN with TTSN had strong robustness and better generalization ability for crater detection within 1 km in different terrains, making it possible to detect small craters with high accuracy when using DOM data.

Correcting data imbalance for semi-supervised COVID-19 detection using X-ray chest images

A key factor in the fight against viral diseases such as the coronavirus (COVID-19) is the identification of virus carriers as early and quickly as possible, in a cheap and efficient manner. The application of deep learning for image classification of chest X-ray images of COVID-19 patients could become a useful pre-diagnostic detection methodology. However, deep learning architectures require large labelled datasets. This is often a limitation when the subject of research is relatively new as in the case of the virus outbreak, where dealing with small labelled datasets is a challenge. Moreover, in such context, the datasets are also highly imbalanced, with few observations from positive cases of the new disease. In this work we evaluate the performance of the semi-supervised deep learning architecture known as MixMatch with a very limited number of labelled observations and highly imbalanced labelled datasets. We demonstrate the critical impact of data imbalance to the model’s accuracy. Therefore, we propose a simple approach for correcting data imbalance, by re-weighting each observation in the loss function, giving a higher weight to the observations corresponding to the under-represented class. For unlabelled observations, we use the pseudo and augmented labels calculated by MixMatch to choose the appropriate weight. The proposed method improved classification accuracy by up to 18%, with respect to the non balanced MixMatch algorithm. We tested our proposed approach with several available datasets using 10, 15 and 20 labelled observations, for binary classification (COVID-19 positive and normal cases). For multi-class classification (COVID-19 positive, pneumonia and normal cases), we tested 30, 50, 70 and 90 labelled observations. Additionally, a new dataset is included among the tested datasets, composed of chest X-ray images of Costa Rican adult patients.

Semisupervised dynamic soft sensor based on complementary ensemble empirical mode decomposition and deep learning

Noise, redundancy, and dynamic characteristics in industrial process data have been regarded as the key factors that affect the measurement accuracy of data-driven soft sensors. In this paper, a semi-supervised dynamic soft sensor is proposed to capture the dynamic characteristics of data while removing noise and redundancy within the data, thus ensuring improved accuracy. Complementary ensemble empirical mode decomposition and isometric feature mapping are combined to reduce noise and redundancy. A semi-supervised deep learning model is designed to capture the dynamic characteristics. Compared with traditional soft sensors, the effectiveness and superiority of this method are verified via an experiment using an air preheater of a power boiler. The proposed method achieves the lowest MAE of 0.1745 and the highest correlation coefficient of 0.9969. Compared to methods without data preprocessing, the MAE of the preprocessed soft sensor decreases by 22.28% on average, while the correlation coefficient increases by 0.24% on average.

Increasing the accuracy of single sequence prediction methods using a deep semi-supervised learning framework.

MotivationOver the past 50 years, our ability to model protein sequences with evolutionary information has progressed in leaps and bounds. However, even with the latest deep learning methods, the modelling of a critically important class of proteins, single orphan sequences, remains unsolved.ResultsBy taking a bioinformatics approach to semi-supervised machine learning, we develop Profile Augmentation of Single Sequences (PASS), a simple but powerful framework for building accurate single-sequence methods. To demonstrate the effectiveness of PASS we apply it to the mature field of secondary structure prediction. In doing so we develop S4PRED, the successor to the open-source PSIPRED-Single method, which achieves an unprecedented Q3 score of 75.3% on the standard CB513 test. PASS provides a blueprint for the development of a new generation of predictive methods, advancing our ability to model individual protein sequences.Availability and implementationThe S4PRED model is available as open source software on the PSIPRED GitHub repository (https://github.com/psipred/s4pred), along with documentation. It will also be provided as a part of the PSIPRED web service (http://bioinf.cs.ucl.ac.uk/psipred/).Supplementary information Supplementary data are available at Bioinformatics online.

Building Extraction from Very-High-Resolution Remote Sensing Images Using Semi-Supervised Semantic Edge Detection

The automated detection of buildings in remote sensing images enables understanding the distribution information of buildings, which is indispensable for many geographic and social applications, such as urban planning, change monitoring and population estimation. The performance of deep learning in images often depends on a large number of manually labeled samples, the production of which is time-consuming and expensive. Thus, this study focuses on reducing the number of labeled samples used and proposing a semi-supervised deep learning approach based on an edge detection network (SDLED), which is the first to introduce semi-supervised learning to the edge detection neural network for extracting building roof boundaries from high-resolution remote sensing images. This approach uses a small number of labeled samples and abundant unlabeled images for joint training. An expert-level semantic edge segmentation model is trained based on labeled samples, which guides unlabeled images to generate pseudo-labels automatically. The inaccurate label sets and manually labeled samples are used to update the semantic edge model together. Particularly, we modified the semantic segmentation network D-LinkNet to obtain high-quality pseudo-labels. Specifically, the main network architecture of D-LinkNet is retained while the multi-scale fusion is added in its second half to improve its performance on edge detection. The SDLED was tested on high-spatial-resolution remote sensing images taken from Google Earth. Results show that the SDLED performs better than the fully supervised method. Moreover, when the trained models were used to predict buildings in the neighboring counties, our approach was superior to the supervised way, with line IoU improvement of at least 6.47% and F1 score improvement of at least 7.49%.

Detection and recognition of ultrasound breast nodules based on semi-supervised deep learning: a powerful alternative strategy.

The successful recognition of benign and malignant breast nodules using ultrasound images is based mainly on supervised learning that requires a large number of labeled images. However, because high-quality labeling is expensive and time-consuming, we hypothesized that semi-supervised learning could provide a low-cost and powerful alternative approach. This study aimed to develop an accurate semi-supervised recognition method and compared its performance with supervised methods and sonographers. The faster region-based convolutional neural network was used for nodule detection from ultrasound images. A semi-supervised classifier based on the mean teacher model was proposed to recognize benign and malignant nodule images. The general performance of the proposed method on two datasets (8,966 nodules) was reported. The detection accuracy was 0.88±0.03 and 0.86±0.02, respectively, on two testing sets (1,350 and 2,220 nodules). When 800 labeled training nodules were available, the proposed semi-supervised model plus 4,396 unlabeled nodules performed better than the supervised learning model (area under the curve (AUC): 0.934±0.026 vs. 0.83±0.050; 0.916±0.022 vs. 0.815±0.049). The performance of the semi-supervised model trained on 800 labeled and 4,396 unlabeled nodules was close to that of the supervised learning model trained on a massive number of labeled nodules (n=5,196) (AUC: 0.934±0.026 vs. 0.952±0.027; 0.916±0.022 vs. 0.918±0.017). Moreover, the semi-supervised model was better than the average accuracy of five human sonographers (AUC: 0.922 vs. 0.889). The semi-supervised model can achieve excellent performance for nodule recognition and be useful for medical sciences. The method reduced the number of labeled images required for training, thus significantly alleviating the difficulty in data preparation of medical artificial intelligence.

A semi-supervised deep learning approach for predicting the functional effects of genomic non-coding variations

BackgroundUnderstanding the functional effects of non-coding variants is important as they are often associated with gene-expression alteration and disease development. Over the past few years, many computational tools have been developed to predict their functional impact. However, the intrinsic difficulty in dealing with the scarcity of data leads to the necessity to further improve the algorithms. In this work, we propose a novel method, employing a semi-supervised deep-learning model with pseudo labels, which takes advantage of learning from both experimentally annotated and unannotated data.ResultsWe prepared known functional non-coding variants with histone marks, DNA accessibility, and sequence context in GM12878, HepG2, and K562 cell lines. Applying our method to the dataset demonstrated its outstanding performance, compared with that of existing tools. Our results also indicated that the semi-supervised model with pseudo labels achieves higher predictive performance than the supervised model without pseudo labels. Interestingly, a model trained with the data in a certain cell line is unlikely to succeed in other cell lines, which implies the cell-type-specific nature of the non-coding variants. Remarkably, we found that DNA accessibility significantly contributes to the functional consequence of variants, which suggests the importance of open chromatin conformation prior to establishing the interaction of non-coding variants with gene regulation.ConclusionsThe semi-supervised deep learning model coupled with pseudo labeling has advantages in studying with limited datasets, which is not unusual in biology. Our study provides an effective approach in finding non-coding mutations potentially associated with various biological phenomena, including human diseases.

A survey on deep learning models for wireless capsule endoscopy image analysis

Abdomen Bleeding, Ulcer, Tumour, Crohn's disease, Celiac disease and other diseases in the gastrointestinal tract (GI) are difficult to diagnose, due to the inescapable inherent difficulty in accessing a volute setting in the human body. Wireless Capsule Endoscopy (WCE) offers a patient-cordial, non-invasive and painless investigation in the GI tract. Automatic detection of anomalies in WCE images using Deep Learning Models improves the detection accuracy but it requires a huge number of labeled data for model training. But these deep models suffer from explain-ability and fail to include expert knowledge in the model decision-making process. By keeping these aspects in mind, this survey aims to identify the opportunities for using Semi-Supervised deep learning models over supervised deep learning methods in Wireless Capsule Endoscopy (WCE) anomaly detection and classification. This paper presents a comprehensive survey on various deep learning solutions for anomaly detection and localization techniques utilized in WCE images in the aspect of performance, complexity, and the quality of the dataset. The survey outlined the proposed Attention and Domain Assisted Generative Adversarial Network (ADA-GAN) based Semi-Supervised Model for WCE anomaly image classification along with initial results. The result derives the shortcomings of the current literature methods and paves the potential research opportunities in the Semi-Supervised models in Wireless capsule endoscopy image analysis.