False Detection Research Articles

Early stage tumor segmentation in breast MRI using shape enhanced U-Net

Breast cancer is one of the most common cancers in women. Early diagnosis using MRI can significantly increase the cure rate. However, breast tumor in early stage is small whose edge is ambiguous, leading to false detection and missing detection. To this end, we propose a shape enhanced U-Net with Transformer encoder layer (called SETE-UNet) for automatically segmenting the tumor from breast cancer. First, Transformer encoder layer is introduced to provide global self-attention mechanisms, achieving obvious promotion for tumor segmentation. Second, a discriminative representation is calculated by the proposed shape enhanced branch, which introduces a novel branch to learn a shape boundary with additional supervision. This proposed module can help our algorithm focus on the relevant boundary and also achieve further performance improvement for tiny tumor. Furthermore, we construct an MRI dataset of breast tumors in early stage for breast tumor segmentation, containing 260 cases, including 243 women and 17 men. We compare the proposed method with the popular algorithms on our dataset of breast cancer in early stage, which shows that our method presents excellent segmentation results on breast MRI containing various organs with respect to MIoU, Precision, Accuracy, Dice, Sensitivity, Specificity, AUC, Hausdorff Distance.

Underwater Rescue Target Detection Based on Acoustic Images.

In order to effectively respond to floods and water emergencies that result in the drowning of missing persons, timely and effective search and rescue is a very critical step in underwater rescue. Due to the complex underwater environment and low visibility, unmanned underwater vehicles (UUVs) with sonar are more efficient than traditional manual search and rescue methods to conduct active searches using deep learning algorithms. In this paper, we constructed a sound-based rescue target dataset that encompasses both the source and target domains using deep transfer learning techniques. For the underwater acoustic rescue target detection of small targets, which lack image feature accuracy, this paper proposes a two-branch convolution module and improves the YOLOv5s algorithm model to design an acoustic rescue small target detection algorithm model. For an underwater rescue target dataset based on acoustic images with a small sample acoustic dataset, a direct fine-tuning using optical image pre-training lacks cross-domain adaptability due to the different statistical properties of optical and acoustic images. This paper therefore proposes a heterogeneous information hierarchical migration learning method. For the false detection of acoustic rescue targets in a complex underwater background, the network layer is frozen during the hierarchical migration of heterogeneous information to improve the detection accuracy. In addition, in order to be more applicable to the embedded devices carried by underwater UAVs, an underwater acoustic rescue target detection algorithm based on ShuffleNetv2 is proposed to improve the two-branch convolutional module and the backbone network of YOLOv5s algorithm, and to create a lightweight model based on hierarchical migration of heterogeneous information. Through extensive comparative experiments conducted on various acoustic images, we have thoroughly validated the feasibility and effectiveness of our method. Our approach has demonstrated state-of-the-art performance in underwater search and rescue target detection tasks.

Discriminative Action Snippet Propagation Network for Weakly Supervised Temporal Action Localization

Weakly supervised temporal action localization (WTAL) aims to classify and localize actions in untrimmed videos with only video-level labels. Recent studies have attempted to obtain more accurate temporal boundaries by exploiting latent action instances in ambiguous snippets or propagating representative action features. However, empirically handcrafted ambiguous snippet extraction and the imprecise alignment of representative snippet propagation lead to challenges in modeling the completeness of actions for these methods. In this article, we propose a Discriminative Action Snippet Propagation Network (DASP-Net) to accurately discover ambiguous snippets in videos and propagate discriminative instance-level features throughout the video for improving action completeness. Specifically, we introduce a novel discriminative feature propagation module for capturing the global contextual attention and propagating the action concept across the whole video by perceiving the discriminative action snippets with instance information from the same video. Simultaneously, we incorporate denoised pseudo-labels as supervision, where we correct the controversial prediction based on the feature space distribution during training, thereby alleviating false detection caused by noise background features. Furthermore, we design an ambiguous feature mining module, which maximizes the feature affinity information of action and background in ambiguous snippets to generate more accurate latent action and background snippets and learns more precise action instance boundaries through contrastive learning of action and background snippets. Extensive experiments show that DASP-Net achieves state-of-the-art results on THUMOS14 and ActivityNet1.2 datasets.

Detection of Ulcerative Colitis Lesions from Weakly Annotated Colonoscopy Videos Using Bounding Boxes

Ulcerative colitis is a chronic disease characterized by bleeding and ulcers in the colon. Disease severity assessment via colonoscopy videos is time-consuming and only focuses on the most severe lesions. Automated detection methods enable fine-grained assessment but depend on the training set quality. To suit the local clinical setup, an internal training dataset containing only rough bounding box annotations around lesions was utilized. Following previous works, we propose to use linear models in suitable color spaces to detect lesions. We introduce an efficient sampling scheme for exploring the set of linear classifiers and removing trivial models i.e., those showing zero false negative or positive ratios. Bounding boxes lead to exaggerated false detection ratios due to mislabeled pixels, especially in the corners, resulting in decreased model accuracy. Therefore, we propose to evaluate the model sensitivity on the annotation level instead of the pixel level. Our sampling strategy can eliminate up to 25% of trivial models. Despite the limited quality of annotations, the detectors achieved better performance in comparison with the state-of-the-art methods. When tested on a small subset of endoscopic images, the best models exhibit low variability. However, the inter-patient model performance was variable suggesting that appearance normalization is critical in this context.

Noise-like signal detector, reducing the probability of false alarms and not giving false detections associated with the correlation properties of sequences modulating the phase of the signal

Noise-like signal detector, reducing the probability of false alarms and not giving false detections associated with the correlation properties of sequences modulating the phase of the signal

A Multi-Source Data Fusion Network for Wood Surface Broken Defect Segmentation.

Wood surface broken defects seriously damage the structure of wooden products, these defects have to be detected and eliminated. However, current defect detection methods based on machine vision have difficulty distinguishing the interference, similar to the broken defects, such as stains and mineral lines, and can result in frequent false detections. To address this issue, a multi-source data fusion network based on U-Net is proposed for wood broken defect detection, combining image and depth data, to suppress the interference and achieve complete segmentation of the defects. To efficiently extract various semantic information of defects, an improved ResNet34 is designed to, respectively, generate multi-level features of the image and depth data, in which the depthwise separable convolution (DSC) and dilated convolution (DC) are introduced to decrease the computational expense and feature redundancy. To take full advantages of two types of data, an adaptive interacting fusion module (AIF) is designed to adaptively integrate them, thereby generating accurate feature representation of the broken defects. The experiments demonstrate that the multi-source data fusion network can effectively improve the detection accuracy of wood broken defects and reduce the false detections of interference, such as stains and mineral lines.

Improved Elman Deep Learning Model for Intrusion Detection System in Internet of Things

Many researchers have developed intrusion detection systems in the past using conventional techniques such as artificial neural networks, fuzzy clustering, evolutionary algorithms, association rule mining, and support vector machines. However, in terms of false negative rates and detection rates, these methods did not yield the best outcomes. To address these problems, we proposed a hybrid deep learning model (HDLM) based on intrusion detection and prevention in IoT devices. Initially, the data are collected from KDDCup-99 and NSL-KDD datasets. Then, the important features are extracted from the dataset using the Forward Feature Selection Algorithm (FFSA). Finally, the extracted features are given to the HDLM classifier. The proposed HDLM is a combination of Elman Recurrent Neural Network (ERNN) and Subtraction-Average-Based Optimizer (SABO). The performance of the suggested method is assessed using performance metrics including precision, recall, accuracy, sensitivity, specificity, and F_Measure. The experimental results show that the proposed method attained the maximum intrusion detection accuracy of 98.52%.

The causality measure of partial mutual information from mixed embedding (PMIME) revisited.

The measure of partial mutual information from mixed embedding (PMIME) is an information theory-based measure to accurately identify the direct and directional coupling, termed Granger causality or simply causality, between the observed variables or subsystems of a high-dimensional dynamical and complex system, without any a priori assumptions about the nature of the coupling relationship. In its core, it is a forward selection procedure that aims to iteratively identify the lag-dependence structure of a given observed variable (response) to all the other observed variables (candidate drivers). This model-free approach is capable of detecting nonlinear interactions, abundantly present in real-world complex systems, and it was shown to perform well on multivariate time series of moderately high dimension. However, the PMIME presents some inefficiencies in its performance mainly when applied on strongly stochastic (linear or nonlinear) systems as it may falsely detect non-existent relationships. Moreover, and by construction, the measure cannot extract purely synergetic relationships present in a system. In the current work, the issue of false detections is addressed by introducing an improved resampling significance test and a procedure of rechecking the identified drivers (backward revision). Regarding the inability to detect synergetic relationships, the PMIME is further enhanced by checking pairs as candidate drivers for the response variable after having considered all drivers individually. The effects of these modifications are investigated in a systematic simulation study on properly designed systems involving strong stochasticity, regressor terms with synergetic effects, and a system dimension ranging from 3 to 30. The overall results of the simulations indicate that these modifications indeed improve the performance of PMIME and alleviate to a significant degree the issues of the original algorithm. Guidelines for balancing between accuracy and computational efficiency are also given, particularly relevant for real-world applications. Finally, the measure performance is investigated in the study of futures of various government bonds and stock market indices in the period around COVID-19 pandemic.

Missing Argo Float Profiles in Highly Stratified Waters of the Amazon River Plume

Abstract In the western tropical Atlantic Ocean close to the Amazon plume, a large loss rate of Argo-float profiles took place, that is, instances of profiles that should have happened but were not transmitted. We find that APEX and SOLO floats were not ascending to the surface in the presence of low surface practical salinity, typically on the order of 32.5 or less, because of limitations on the surface buoyancy range for those floats. This results in an overall loss of profiles from these floats that is on the order of 6% averaged over the year, with a peak of 12% in July. We also find aborted descents/incorrect grounding detections for ARVOR/PROVOR floats when surface salinity is low and the descending float reaches a strong halocline (2.6% of all the profiles in the June–August season). Altogether, the whole Argo set includes a maximum loss rate of roughly 6% in July. We find a pattern of loss that fits the surface salinity seasonal cycle and the occurrence of low surface salinity investigated from a high-resolution daily satellite salinity product in 2010–21. The agreement is even better when considering surface density instead of surface salinity, with the temperature contribution to density inducing a shift in the maximum occurrence of these events by 1 month relative to the cycle of very low salinity events. Because of changes in the float technology, the loss rate that targets the lowest surface salinities was very large until 2010, with an overall decrease afterward. Significance Statement In the western tropical Atlantic Ocean, some Argo floats were not able to ascend or descend with very low surface salinity, because of buoyancy limitations for some float types and false bottom detection on others. In this region, for surface practical salinity smaller than 32.5, this resulted in a loss of close to one-half of the Argo profiles during the last 20 years. Altogether, this undersampling of the lowest surface salinities by Argo floats modifies the upper-ocean salinity seasonal cycle, as well as longer-term trends portrayed in Argo data–based products. Furthermore, in this region, care must be taken when validating satellite salinity data with Argo data or when adjusting satellite sea surface salinity data to in situ data products.

Deep self-taught learning framework for intrusion detection in cloud computing environment

<p>Cloud has become a target-rich environment for malicious attacks by cyber intruders. Security is a major concern and remains an obstacle to the adoption of cloud computing. The intrusion detection system (IDS) is regarded as defense-in-depth. Unfortunately, most machine learning approaches designed for cloud intrusion detection require large amounts of labeled attack samples, but in real practice, they are limited. Therefore, the key impetus of this work is to introduce self-taught learning (STL) combining stacked sparse autoencoder (SSAE) with long short-term memory (LSTM) as a candidate solution to learn the robust feature representation and efficiently improve the performance of IDS with respect to false alarm rate (FAR) and detection rate (DR). Accordingly, the proposed approach as a first step employs SSAE to achieve dimensional reduction by learning the discriminative features from network traffic. The approach adopts LSTM to recognize the intrusion with the features encoded by SSAE. To evaluate the detective performance of our model, a comprehensive set of experiments are conducted on NSL-KDD. Also, ablation experiments are conducted to show the contribution of each component of our approach. Further, the comparative analysis shows the efficacy of our approach against the existing approaches with an accuracy of 86.31%.</p>

High‐volume plankton tow net sampling improves eDNA detection of invasive zebra mussels (Dreissena polymorpha) in recently infested lakes

AbstractAquatic invasive species are a serious and growing threat to biodiversity. Zebra and quagga mussels (Dreissena polymorpha and D. rostriformis bugensis) are freshwater invaders causing substantial ecological and economic damage across Europe and North America. Early detection of invasive mussels and other non‐indigenous species is increasingly needed to prevent their establishment and spread. Environmental DNA (eDNA) techniques potentially offer higher sensitivity monitoring tools to complement more conventional methods for surveying adult and juvenile mussels and veliger larvae. eDNA assays are typically performed on small‐volume (0.5–5 L) water samples that are concentrated by filtration prior to extraction and downstream processing. Sampling using a towed plankton net of a larger (64 μm) pore size can process orders of magnitude larger water volumes with the potential for increasing eDNA detection sensitivity. We compared the sensitivities of high‐volume plankton tow net water sampling to filter sampling in three recently infested lakes in Canada and Minnesota, USA. Paired filtration and tow net samples were analyzed for Dreissena DNA using an established quantitative polymerase chain reaction assay for the genus. Higher yields of Dreissena eDNA (more DNA copies) were recovered from plankton tow than from filtered samples in all 33 paired comparisons. In some cases, plankton tow samples were positive for Dreissena eDNA while lower‐volume filtering produced a false negative detection. Our results demonstrate the effectiveness of plankton tow net sampling for eDNA early detection of invasive mussels, a method that can be used exclusively or as a supplement to filter sampling. Our results further suggest that eDNA testing could be incorporated into monitoring programs that routinely use plankton tows for visual detection of invasive mollusk larvae, as well as other aquatic invasive and non‐invasive species ranging from plankton to metazoans, including many fish.

Detection of Electrically Evoked Auditory Steady-State Responses in Cochlear-Implant Recipients With a System Identification Based Method.

Electrically evoked auditory steady-state responses (EASSRs) can potentially be used as an objective measure to realize the automatic fitting of cochlear implants (CIs). They can be recorded using electroencephalography (EEG) and objectively detected at the modulation frequency of the stimulus. The main roadblock in using EASSRs is the presence of CI stimulation artifacts in the EEG recording. In this article, we present an improvement of a recently introduced system identification (SI) based artifact removal method. We evaluate its applicability for objective CI fitting on a larger dataset. The parameter estimation problem of the SI is solved using ordinary least squares (OLS), where an additional regularization term is added to the cost function. We compare EASSR latencies as determined by the commonly used linear interpolation artifact removal method and SI, to evaluate the artifact removal and EASSR detection quality on a dataset of 16 CI recipients and four different stimulation levels. SI can fully remove stimulation artifacts and detect EASSRs, even for recordings from ipsilateral EEG channels, where all other artifact removal methods fail so far. Using OLS with regularization prevents false positive response detection. Using SI, EASSRs can reliably be detected in EEG recordings, even for ipsilateral recording channels and recordings with lower stimulation levels. As the recordings are obtained with clinically relevant settings of the CI, they reveal the potential impact of SI on the objective fitting of CIs. We argue, that SI enables therefore a big step towards automated CI fitting with EASSRs.

Enabling Efficient and Malicious Secure Data Aggregation in Smart Grid With False Data Detection

Enabling Efficient and Malicious Secure Data Aggregation in Smart Grid With False Data Detection

Two-Step Accuracy Improvement for Multitarget Detection in Complex Environment Using UWB Radar

Detecting multiple human targets in indoor scenarios using ultra-wideband (UWB) radar usually involves false detection results caused by the secondary reflections, which might reduce the target detection accuracy and cause a more severe deterioration when the number of targets increases. This article proposed a two-step accuracy improvement method for multitarget detection in environments with multiple human targets of more than three and strong secondary reflections by the surroundings, especially the walls. Based on the rough detection results acquired by the modified CA-CFAR (MCA-CFAR) processing, the first step achieves the primary false alarm suppression using a short-window accumulation in the time domain. Then, the second step applies the decision confidence on the detection results from the first step to assess the reliability of results for improved accuracy. The two-step accuracy improvement could thus have a higher accuracy through cascading false alarm suppression. The effectiveness and accuracy of the proposed algorithm are verified based on the experimental results.

FAKE NEWS DETECTION USING MACHINE LEARNING MULTI-MODEL METHOD

A fake news article that originates from an WhatsApp source is known as fake news. Fake news is becoming more and more prevalent on social media and other platforms, and this is a serious worry since it has the potential to have devastating effects on society and the country. This is why there has already been a lot of research done on its detection. This study uses supervised machine learning techniques to develop a product model through research and implementation of false news detection system. To put it briefly, this work will use a Naive Bayes classifier to build a model that can identify fake news by measuring its words and phrases against a set of criteria. that uses techniques like a count vectorizer (using word tallies) or a (Term Frequency Inverse Document Frequency) tfidf matrix to categorize bogus news as real or false It's highly likely that the meaning of two papers with comparable word counts will be entirely different.

Semiparametric constant false alarm rate method for radar and sonar images

AbstractThis work proposed a novel constant false alarm rate detection method based on Gaussian mixture model (GMM). The main contribution contains two aspects. First, the Gabor wavelet is used to determine the GMM component number. Second, the expression of probability of false alarm is derived based on the GMM parameter.

GD3N: Adaptive clustering-based detection of selective forwarding attacks in WSNs under variable harsh environments

Wireless sensor networks (WSNs) are susceptible to numerous security threats due to their reliance on open environments and broadcast communication methods. Among these, the selective forwarding attack is notably challenging to detect. This difficulty arises from the ability of malicious nodes to imitate the behavior of normal nodes, and selectively drop data packets, which makes them virtually indistinguishable from normal ones, particularly under conditions of poor channel quality. To address this challenge with harsh environments, we introduce a novel methodology termed GD3N. This approach is underpinned by the design of a unique type of data point that encapsulates both short-term and long-term forwarding behaviors of nodes. It combines a refined version of the Gradient Diffusion Density-based Spatial Clustering of Applications with Noise (GD-DBSCAN) algorithm, with a novel Double-Parameter Neighbor Voting (DP-NV) method based on the data set. These innovations contribute to a significant enhancement in detection accuracy and a reduction in computational complexity when compared to traditional DBSCAN and NV methods. Simulation results show that our GD3N achieves a false detection rate (FDR) of less than 2%, a missed detection rate (MDR) of below 10%, and an overall detection accuracy rate (DAR) of over 95% across various testing scenarios.

Defining a metagenomic threshold for detecting low abundances of Providencia alcalifaciens in canine faecal samples.

Acute haemorrhagic diarrhoea syndrome (AHDS) in dogs is a condition of unknown aetiology. Providencia alcalifaciens is suspected to play a role in the disease as it was commonly found in dogs suffering from AHDS during a Norwegian outbreak in 2019. The role of this bacterium as a constituent of the canine gut microbiota is unknown, hence this study set out to investigate its occurrence in healthy dogs using metagenomics. To decrease the likelihood of false detection, we established a metagenomic threshold for P. alcalifaciens by spiking culture-negative stool samples with a range of bacterial dilutions and analysing these by qPCR and shotgun metagenomics. The detection limit for P. alcalifaciens was determined and used to establish a metagenomic threshold. The threshold was validated on naturally contaminated faecal samples with known cultivation status for P. alcalifaciens. Finally, the metagenomic threshold was used to determine the occurrence of P. alcalifaciens in shotgun metagenomic datasets from canine faecal samples (n=362) collected in the HUNT One Health project. The metagenomic assay and qPCR had a detection limit of 1.1x103 CFU P. alcalifaciens per faecal sample, which corresponded to a Cq value of 31.4 and 569 unique k-mer counts by shotgun metagenomics. Applying this metagenomic threshold to 362 faecal metagenomic datasets from healthy dogs, P. alcalifaciens was found in only 1.1% (95% CI [0.0, 6.8]) of the samples, and then in low relative abundances (median: 0.04%; range: 0.00 to 0.81%). The sensitivity of the qPCR and shotgun metagenomics assay was low, as only 40% of culture-positive samples were also positive by qPCR and metagenomics. Using our detection limit, the occurrence of P. alcalifaciens in faecal samples from healthy dogs was low. Given the low sensitivity of the metagenomic assay, these results do not rule out a significantly higher occurrence of this bacterium at a lower abundance.

A single-stage anchor-free rotating target visual detection algorithm suitable for flexible body vibration displacement measurement

The phenomenon of angular inclination of flexible structures during vibration poses a significant challenge to the applicability of visual vibration measurement methods because the target locked in the captured image will produce unknown geometric deformations such as scale, displacement, and angle in the time domain space, and the horizontal rectangular frame used for matching during target detection will also increase the false detection rate of the target due to the introduction of more background information. Such subtle geometric deformations and false detections can lead to severe fit errors in the displacement curves regressed by the visual vibration measurement algorithm. To effectively improve the accuracy and robustness of vibration image target recognition, this article takes the flexible body captured by a high-speed camera as the target of vibration displacement measurement. It introduces the rotating target detection method based on deep learning into the field of visual vibration measurement, which verifies the feasibility of the deep learning method in flexible body vibration measurement, and based on the deep convolutional neural network framework, a high-precision displacement measurement algorithm based on single-stage anchor-free rotating target detection is proposed. The algorithm in this article first uses the CSPDarknet backbone network to extract multi-scale features of flexible structural image sequences. It then uses PANet to fuse the top-down and bottom-up bidirectional feature maps of the four bridge target feature maps obtained through the backbone network. The shallow and deep information is used for semantic feature fusion and combined with the Coordinate Attention mechanism to achieve target finding and fine positioning on the feature map. Finally, we use the coordinates of the bounding box obtained from the test to regress the position offset of the object’s center point. To verify the accuracy of the algorithm in this article, we conducted experimental validation on the cable-stayed bridge model and the actual bridge and compared the performance with the traditional template matching algorithm, differential optical flow method, and various deep learning algorithms with different localization principles, as well as the displacement signals collected and processed by accelerometers. The experimental results of time-frequency characteristics analysis show that the vibration displacement trajectories regressed by the algorithm in this paper have the best overlap with the displacement measurements collected by the accelerometer, which verifies that the algorithm in this article has good application potential and implementation space in the field of condition monitoring of flexible structural bodies.

Unsupervised Bayesian change-point detection approach for reliable prognostics and health management of complex mechanical systems

In the prognostics and health management of mechanical systems, achieving high reliability requires high-quality data for analysis and model training to avoid intensive misalarms and false detections. This research introduces a novel unsupervised Bayesian change-point detection approach specifically tailored to improve the accuracy and dependability of data used in the performance evaluation, fault diagnostics, and predictive maintenance of such systems. By leveraging this method, the precision of change-point detection in operational data is significantly heightened, effectively discerning between normal conditions and potential outliers within the recorded data sequences. The proposed approach replaces the reliance on traditional outlier detection methods with an innovative utilization of prediction error metrics, enriched with statistical analysis to assess and bolster the robustness of the detection algorithm. Conducted case studies on data from the supervisory control and data acquisition (SCADA) system of 92 wind turbines highlight the advanced data cleaning capabilities of our method, which markedly outperforms previous techniques including quartile, Gaussian mixture models, isolation forests, and local outlier factor algorithms. The findings illustrate the methodology’s efficacy in obtaining rcleaned data, which are crucial for the reliable prognostics and health management frameworks of mechanic system, and enhance the confidence in feature extraction and performance prediction.