Learning-based Solution Research Articles

CC4S: Encouraging Certainty and Consistency in Scribble-Supervised Semantic Segmentation.

Deep learning-based solutions have achieved impressive performance in semantic segmentation but often require large amounts of training data with fine-grained annotations. To alleviate such requisition, a variety of weakly supervised annotation strategies have been proposed, among which scribble supervision is emerging as a popular one due to its user-friendly annotation way. However, the sparsity and diversity of scribble annotations make it nontrivial to train a network to produce deterministic and consistent predictions directly. To address these issues, in this paper we propose holistic solutions involving the design of network structure, loss and training procedure, named CC4S to improve Certainty and Consistency for Scribble-Supervised Semantic Segmentation. Specifically, to reduce uncertainty, CC4S embeds a random walk module into the network structure to make neural representations uniformly distributed within similar semantic regions, which works together with a soft entropy loss function to force the network to produce deterministic predictions. To encourage consistency, CC4S adopts self-supervision training and imposes the consistency loss on the eigenspace of the probability transition matrix in the random walk module (we named neural eigenspace). Such self-supervision inherits the category-level discriminability from the neural eigenspace and meanwhile helps the network focus on producing consistent predictions for the salient parts and neglect semantically heterogeneous backgrounds. Finally, to further improve the performance, CC4S uses the network predictions as pseudo-labels and retrains the network with an extra color constraint regularizer. From comprehensive experiments, CC4S achieves comparable performance to those from fully supervised methods and shows promising robustness under extreme supervision cases.

Deep reinforcement learning-based spatio-temporal graph neural network for solving job shop scheduling problem

AbstractThe job shop scheduling problem (JSSP) is a well-known NP-hard combinatorial optimization problem that focuses on assigning tasks to limited resources while adhering to certain constraints. Currently, deep reinforcement learning (DRL)-based solutions are being widely used to solve the JSSP by defining the problem structure on disjunctive graphs. Some of the proposed approaches attempt to leverage the structural information of the JSSP to capture the dynamics of the environment without considering the time dependency within the JSSP. However, learning graph representations only from the structural relationship of nodes results in a weak and incomplete representation of these graphs which does not provide an expressive representation of the dynamics in the environment. In this study, unlike existing frameworks, we defined the JSSP as a dynamic graph to explicitly consider the time-varying aspect of the JSSP environment. To this end, we propose a novel DRL framework that captures both the spatial and temporal attributes of the JSSP to construct rich and complete graph representations. Our DRL framework introduces a novel attentive graph isomorphism network (Attentive-GIN)-based spatial block to learn the structural relationship and a temporal block to capture the time dependency. Additionally, we designed a gated fusion block that selectively combines the learned representations from the two blocks. We trained the model using the proximal policy optimization algorithm of reinforcement learning. Experimental results show that our trained model exhibits significant performance enhancement compared to heuristic dispatching rules and learning-based solutions for both randomly generated datasets and public benchmarks.

Integrating learning-based solutions in intelligent transportation systems: a conceptual framework and case studies validation

Integrating learning-based solutions in intelligent transportation systems: a conceptual framework and case studies validation

Detection of Fake Instagram Accounts via Machine Learning Techniques

This paper focuses on the detection of fake accounts on Instagram and proposes a novel solution that aims to address this problem. More specifically, a machine learning-based solution is introduced that can be employed by Instagram-based applications to combat this phenomenon. To accomplish this, publicly available data from Instagram users are collected and processed. After making the necessary feature additions to and removals from these data, they are fed into machine learning algorithms with the aim of detecting fake Instagram accounts. The interest of the results presented in this study lies in their comparability with the results of other similar studies, with the difference that fewer input features have been used. This addresses significant research challenges regarding the performance of and confidence in obtained results in this specific problem domain.

Pneumonia detection on chest X-rays from Xception-based transfer learning and logistic regression.

Pneumonia is a dangerous disease that kills millions of children and elderly patients worldwide every year. The detection of pneumonia from a chest x-ray is perpetrated by expert radiologists. The chest x-ray is cheaper and is most often used to diagnose pneumonia. However, chest x-ray-based diagnosis requires expert radiologists which is time-consuming and laborious. Moreover, COVID-19 and pneumonia have similar symptoms which leads to false positives. Machine learning-based solutions have been proposed for the automatic prediction of pneumonia from chest X-rays, however, such approaches lack robustness and high accuracy due to data imbalance and generalization errors. This study focuses on elevating the performance of machine learning models by dealing with data imbalanced problems using data augmentation. Contrary to traditional machine learning models that required hand-crafted features, this study uses transfer learning for automatic feature extraction using Xception and VGG-16 to train classifiers like support vector machine, logistic regression, K nearest neighbor, stochastic gradient descent, extra tree classifier, and gradient boosting machine. Experiments involve the use of hand-crafted features, as well as, transfer learning-based feature extraction for pneumonia detection. Performance comparison using Xception and VGG-16 features suggest that transfer learning-based features tend to show better performance than hand-crafted features and an accuracy of 99.23% can be obtained for pneumonia using chest X-rays.

Using spatial video and deep learning for automated mapping of ground-level context in relief camps

BackgroundThe creation of relief camps following a disaster, conflict or other form of externality often generates additional health problems. The density of people in a highly stressed environment with questionable safe food and water access presents the potential for infectious disease outbreaks. These camps are also not static data events but rather fluctuate in size, composition, and level and quality of service provision. While contextualized geospatial data collection and mapping are vital for understanding the nature of these camps, various challenges, including a lack of data at the required spatial or temporal granularity, as well as the issue of sustainability, can act as major impediments. Here, we present the first steps toward a deep learning-based solution for dynamic mapping using spatial video (SV).MethodsWe trained a convolutional neural network (CNN) model on a SV dataset collected from Goma, Democratic Republic of Congo (DRC) to identify relief camps from video imagery. We developed a spatial filtering approach to tackle the challenges associated with spatially tagging objects such as the accuracy of global positioning system and positioning of camera. The spatial filtering approach generates smooth surfaces of detection, which can further be used to capture changes in microenvironments by applying techniques such as raster math.ResultsThe initial results suggest that our model can detect temporary physical dwellings from SV imagery with a high level of precision, recall, and object localization. The spatial filtering approach helps to identify areas with higher concentrations of camps and the web-based tool helps to explore these areas. The longitudinal analysis based on applying raster math on the detection surfaces revealed locations, which had a considerable change in the distribution of tents over space and time.ConclusionsThe results lay the groundwork for automated mapping of spatial features from imagery data. We anticipate that this work is the building block for a future combination of SV, object identification and automatic mapping that could provide sustainable data generation possibilities for challenging environments such as relief camps or other informal settlements.

Comparative Analysis of Machine Learning Models for Accurate Flight Price Prediction

Flight fare prediction is a vital component in helping consumers make informed decisions regarding travel expenses. Airline ticket prices fluctuate due to a variety of factors such as demand, time of purchase, and flight routes. In this research, we propose a machine learning-based solution for predicting flight fares using historical data. Models like Random Forest, Gradient Boosting, and Support Vector Machines (SVM) are employed to analyze flight data and produce reliable predictions. This study demonstrates how predictive models can benefit customers by offering insights into pricing trends, thus optimizing their flight booking process.

Experts still needed: boosting long-term android malware detection with active learning

The continuous evolution of cyber threats imposes a critical challenge to malware detection systems, so operational detection solutions in real-world settings must keep up-to-date malware knowledge databases. Machine learning-based solutions are not exempt from this requirement as handling concept drift constitutes the primary building block for keeping high detection performance in the long term. However, maintaining non-stationary malware detection models is highly demanding due to the high cost of labeling. This study applies several active learning-based approaches for maintaining a non-stationary model for Android malware detection in a 7-year-long time frame and conducts a comprehensive analysis to understand the impact of feature space selection, different data balancing techniques, and timestamping methods, utilized for locating the instances along the historical timeline, on the model’s detection performance over time. The detection accuracy and labeling costs are compared with various baselines. Additionally, the study investigates the resilience of such models against noisy labeling, a common problem in production environments due to unintentional expert errors and adversarial attacks. This research fills a significant gap in the literature by conducting a comprehensive analysis of active learning approaches to address concept drift in non-stationary settings established for mobile malware detection.

The Use of eXplainable Artificial Intelligence and Machine Learning Operation Principles to Support the Continuous Development of Machine Learning-Based Solutions in Fault Detection and Identification

Machine learning (ML) revolutionized traditional machine fault detection and identification (FDI), as complex-structured models with well-designed unsupervised learning strategies can detect abnormal patterns from abundant data, which significantly reduces the total cost of ownership. However, their opaqueness raised human concern and intrigued the eXplainable artificial intelligence (XAI) concept. Furthermore, the development of ML-based FDI models can be improved fundamentally with machine learning operations (MLOps) guidelines, enhancing reproducibility and operational quality. This study proposes a framework for the continuous development of ML-based FDI solutions, which contains a general structure to simultaneously visualize and check the performance of the ML model while directing the resource-efficient development process. A use case is conducted on sensor data of a hydraulic system with a simple long short-term memory (LSTM) network. Proposed XAI principles and tools supported the model engineering and monitoring, while additional system optimization can be made regarding input data preparation, feature selection, and model usage. Suggested MLOps principles help developers create a minimum viable solution and involve it in a continuous improvement loop. The promising result motivates further adoption of XAI and MLOps while endorsing the generalization of modern ML-based FDI applications with the HITL concept.

DeepMarkerNet: Leveraging supervision from the Duchenne Marker for spontaneous smile recognition

Distinguishing between spontaneous and posed smiles from videos poses a significant challenge in pattern classification literature. Researchers have developed feature-based and deep learning-based solutions for this problem. To this end, deep learning outperforms feature-based methods. However, certain aspects of feature-based methods could improve deep learning methods. For example, previous research has shown that Duchenne Marker (or D-Marker) features from the face play a vital role in spontaneous smiles, which can be useful to improve deep learning performances. In this study, we propose a deep learning solution that leverages D-Marker features to improve performance further. Our multi-task learning framework, named DeepMarkerNet, integrates a transformer network with the utilization of facial D-Markers for accurate smile classification. Unlike past methods, our approach simultaneously predicts the class of the smile and associated facial D-Markers using two different feed-forward neural networks, thus creating a symbiotic relationship that enriches the learning process. The novelty of our approach lies in incorporating supervisory signals from the pre-calculated D-Markers (instead of as input in previous works), harmonizing the loss functions through a weighted average. In this way, our training utilizes the benefits of D-Markers, but the inference does not require computing the D-Marker. We validate our model’s effectiveness on four well-known smile datasets: UvA-NEMO, BBC, MMI facial expression, and SPOS datasets, and achieve state-of-the-art results.

Deep Learning based Counterfeit Nike Shoes Detection using YOLOv8 for Object Detection

The fashion and footwear industries, where brand value and customer trust are paramount, are under constant threat from counterfeit products. This research presents a deep learning-based solution for detecting counterfeit Nike shoes using the YOLOv8 model for object detection. the dataset included four classes: “Nike Fake Air Force,” “Nike Fake Jordan 1,” “Nike Original Air Force,” and “Nike Original Jordan 1,” comprising a total of 3,860 images. These were split into training (70%), validation (20%), and testing (10%) sets. The pre-trained, medium-sized YOLOv8 model was used for detection and classification, yielding promising results. The model achieved a mAP of 95.0%, with precision and recall scores of 92.2% and 91.8%, respectively, on the validation set. The images were web scrapped with the Chrome extension called "Download All Images" and then manually filtered so that they would be relevant and of good quality. Each image was then manually labelled using the RoboFlow platform. Nevertheless, the model appears to be promising for implementation in combating counterfeit products, with a high potential accuracy and efficiency.

Recognition of Malicious URLs Using Machine Learning

The surge in online activities has heightened the risk of cyber-attacks, with phishing emerging as a notably widespread threat. Conventional phishing detection methods, such as blacklists and heuristic analysis, fall short when it comes to identifying novel and complex phishing schemes. To tackle this issue, we propose a machine learning-based solution for detecting phishing websites. Our model utilizes a combination of features—such as URL structure, domain attributes, and content characteristics—to categorize websites as either phishing or legitimate. The model's outstanding overall accuracy of 96.9% underscores the need to evaluate both precision and recall when determining the effectiveness of machine learning models. The results of this study are significant for developing robust URL classification models and advancing the field of cybersecurity. Future research could aim to improve the model's performance by integrating more features, developing real-time phishing detection systems, and exploring new attributes that can be derived from URLs. Keywords: Phishing, URL, Machine Learning, Classification, Detection

NABNet: Deep Learning-Based IoT Alert System for Detection of Abnormal Neck Behavior.

The excessive use of electronic devices for prolonged periods has led to problems such as neck pain and pressure injury in sedentary people. If not detected and corrected early, these issues can cause serious risks to physical health. Detectors for generic objects cannot adequately capture such subtle neck behaviors, resulting in missed detections. In this paper, we explore a deep learning-based solution for detecting abnormal behavior of the neck and propose a model called NABNet that combines object detection based on YOLOv5s with pose estimation based on Lightweight OpenPose. NABNet extracts the detailed behavior characteristics of the neck from global to local and detects abnormal behavior by analyzing the angle of the data. We deployed NABNet on the cloud and edge devices to achieve remote monitoring and abnormal behavior alarms. Finally, we applied the resulting NABNet-based IoT system for abnormal behavior detection in order to evaluate its effectiveness. The experimental results show that our system can effectively detect abnormal neck behavior and raise alarms on the cloud platform, with the highest accuracy reaching 94.13%.

MIMO Signal Detection Based on IM-LSTMNet Model

Signal detection is crucial in multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM) systems, yet classical detection methods often struggle with nonlinear issues in wireless channels. To handle this challenge, we propose a novel signal detection method for MIMO-OFDM system based on the fractional Fourier transform (FrFT), leveraging the robust time series processing capabilities of long short-term memory (LSTM) networks. Our innovative approach, termed IM-LSTMNet, integrates LSTM with convolutional neural networks (CNNs) and incorporates a Squeeze and Excitation Network to emphasize critical information, enhancing neural network performance. The proposed IM-LSTMNet is applied to the FrFT-based MIMO-OFDM system to improve signal detection performance. We compare the detection results of IM-LSTMNet with zero forcing (ZF), minimum mean square error (MMSE), simple LSTM neural network, and CNN–LSTM network by evaluating the bit error rate. Experimental results demonstrate that IM-LSTMNet outperforms ZF, MMSE, LSTM, and other methods, significantly enhancing system signal detection performance. This work offers a promising advancement in MIMO-OFDM signal detection, presenting a deep learning-based solution that effectively improves the system signal detection performance.

Ground-based image deconvolution with Swin Transformer UNet

Aims. As ground-based all-sky astronomical surveys will gather millions of images in the coming years, a critical requirement emerges for the development of fast deconvolution algorithms capable of efficiently improving the spatial resolution of these images. By successfully recovering clean and high-resolution images from these surveys, the objective is to deepen the understanding of galaxy formation and evolution through accurate photometric measurements. Methods. We introduce a two-step deconvolution framework using a Swin Transformer architecture. Our study reveals that the deep learning-based solution introduces a bias, constraining the scope of scientific analysis. To address this limitation, we propose a novel third step relying on the active coefficients in the sparsity wavelet framework. Results. We conducted a performance comparison between our deep learning-based method and Firedec, a classical deconvolution algorithm, based on an analysis of a subset of the EDisCS cluster samples. We demonstrate the advantage of our method in terms of resolution recovery, generalisation to different noise properties, and computational efficiency. The analysis of this cluster sample not only allowed us to assess the efficiency of our method, but it also enabled us to quantify the number of clumps within these galaxies in relation to their disc colour. This robust technique that we propose holds promise for identifying structures in the distant universe through ground-based images.

Synthetic data for learning-based knowledge discovery

Recent advances in deep learning have demonstrated the ability of learning-based methods to tackle very hard downstream tasks. Historically, this has been demonstrated in predictive tasks, while tasks more akin to the traditional KDD (Knowledge Discovery in Databases) pipeline have enjoyed proportionally fewer advances. Can learning-based approaches help with inherently hard problems within the KDD pipeline, such as "how many patterns are in the data", "what are different structures in the data", and "how can we robustly extract those structures?" In this vision paper, we argue for the need for synthetic data generators to empower cheaply-supervised learning-based solutions for knowledge discovery. We describe the general idea, early proof-of-concept results which speak to the viability of the paradigm, and we outline a number of exciting challenges that await, and a set of milestones for measuring success.

ChickenSense: A Low-Cost Deep Learning-Based Solution for Poultry Feed Consumption Monitoring Using Sound Technology

This research proposes a low-cost system consisting of a hardware setup and a deep learning-based model to estimate broiler chickens’ feed intake, utilizing audio signals captured by piezoelectric sensors. The signals were recorded 24/7 for 19 consecutive days. A subset of the raw data was chosen, and events were labeled in two classes, feed-pecking and non-pecking (including singing, anomaly, and silence samples). Next, the labeled data were preprocessed through a noise removal algorithm and a band-pass filter. Then, the spectrogram and the signal envelope were extracted from each signal and fed as inputs to a VGG-16-based convolutional neural network (CNN) with two branches for 1D and 2D feature extraction followed by a binary classification head to classify feed-pecking and non-pecking events. The model achieved 92% accuracy in feed-pecking vs. non-pecking events classification with an f1-score of 91%. Finally, the entire raw dataset was processed utilizing the developed model, and the resulting feed intake estimation was compared with the ground truth data from scale measures. The estimated feed consumption showed an 8 ± 7% mean percent error on daily feed intake estimation with a 71% R2 score and 85% Pearson product moment correlation coefficient (PPMCC) on hourly intake estimation. The results demonstrate that the proposed system estimates broiler feed intake at each feeder and has the potential to be implemented in commercial farms.

A survey for table recognition based on deep learning

Tables, as an important information interaction media containing much structured information, are popularly adopted in our life. There are a lot of table recognition methods for understanding and applying these information. Among them, table recognition based on deep learning is the mainstream. Therefore, we conducted this survey to summarize the latest research on table recognition based on deep learning in recent years. Table recognition can be divided into three sub-tasks, namely table detection, table structure recognition and table content recognition, so we firstly introduce the importance and significance of table recognition research and analyze the current challenges faced by each sub-task. Then, we list the existing public table recognition datasets and introduce commonly used evaluation metrics. Furthermore, we review deep learning-based solutions proposed in recent years for three different sub-tasks and reclassify them. Especially for table structure recognition, we reclassify them according to their annotation schemes. Finally, we report the performance of some excellent methods, such as CascadeTabNet, TSRFormer, SEM, etc, on public datasets, and describe the future research directions, including table recognition in natural scenes, question and answer about tables, visual information extraction, domain adaptation etc. In summary, this survey reviews the data, evaluations, methods, and development directions of table recognition. Table understanding and domain adaptation need as well to be focused on. The former can predict the relationships between two table elements, which helps us to understand the table. The latter can enhance the generalization ability of table recognition methods, thereby obtaining more robust recognition performance. Both are important development directions for table recognition in the future.

Pixel-wise segmentation of cells in digitized Pap smear images

A simple and cheap way to recognize cervical cancer is using light microscopic analysis of Pap smear images. Training artificial intelligence-based systems becomes possible in this domain, e.g., to follow the European recommendation to screen negative smears to reduce false negative cases. The first step for such a process is segmenting the cells. A large and manually segmented dataset is required for this task, which can be used to train deep learning-based solutions. We describe a corresponding dataset with accurate manual segmentations for the enclosed cells. Altogether, the APACS23 (Annotated PAp smear images for Cell Segmentation 2023) dataset contains about 37 000 manually segmented cells and is separated into dedicated training and test parts, which could be used for an official benchmark of scientific investigations or a grand challenge.

Adaptive Multi-Dimensional Weighted Network With Category-Aware Contrastive Learning for Fine-Grained Hand Bone Segmentation.

Accurately delineating and categorizing individual hand bones in 3D ultrasound (US) is a promising technology for precise digital diagnostic analysis. However, this is a challenging task due to the inherent imaging limitations of the US and the insignificant feature differences among numerous bones. In this study, we have proposed a novel deep learning-based solution for pediatric hand bone segmentation in the US. Our method is unique in that it allows for effective detailed feature mining through an adaptive multi-dimensional weighting attention mechanism. It innovatively implements a category-aware contrastive learning method to highlight inter-class semantic feature differences, thereby enhancing the category discrimination performance of the model. Extensive experiments on the challenging pediatric clinical hand 3D US datasets show the outstanding performance of the proposed method in segmenting thirty-eight bone structures, with the average Dice coefficient of 90.0%. The results outperform other state-of-the-art methods, demonstrating its effectiveness in fine-grained hand bone segmentation. Our method will be globally released as a plugin in the 3D Slicer, providing an innovative and reliable tool for relevant clinical applications.