scholarly journals An Enhanced Multimodal Stacking Scheme for Online Pornographic Content Detection

2020 ◽  
Vol 10 (8) ◽  
pp. 2943
Author(s):  
Kwangho Song ◽  
Yoo-Sung Kim
Keyword(s):  
Average Rate ◽  
False Negative ◽  
False Negative Rate ◽  
Serial Order ◽  
True Positive Rate ◽  
Online Detection ◽  
Convolutional Network ◽  
Signal Change ◽  
Positive Rate ◽  
Auditory Features

An enhanced multimodal stacking scheme is proposed for quick and accurate online detection of harmful pornographic contents on the Internet. To accurately detect harmful contents, the implicative visual features (auditory features) are extracted using a bi-directional RNN (recurrent neural network) with VGG-16 (a multilayered dilated convolutional network) to implicitly express the signal change patterns over time within each input. Using only the implicative visual and auditory features, a video classifier and an audio classifier are trained, respectively. By using both features together, one fusion classifier is also trained. Then, these three component classifiers are stacked in the enhanced ensemble scheme to reduce the false negative errors in a serial order of the fusion classifier, video classifier, and audio classifier for a quick online detection. The proposed multimodal stacking scheme yields an improved true positive rate of 95.40% and a false negative rate of 4.60%, which are superior values to previous studies. In addition, the proposed stacking scheme can accurately detect harmful contents up to 74.58% and an average rate of 62.16% faster than the previous stacking scheme. Therefore, the proposed enhanced multimodal stacking scheme can be used to quickly and accurately filter out harmful contents in the online environments.

scholarly journals Pretherapeutic Imaging for Axillary Staging in Breast Cancer: A Systematic Review and Meta-Analysis of Ultrasound, MRI and FDG PET

2021 ◽  
Vol 10 (7) ◽  
pp. 1543
Author(s):  
Morwenn Le Boulc’h ◽  
Julia Gilhodes ◽  
Zara Steinmeyer ◽  
Sébastien Molière ◽  
Carole Mathelin
Keyword(s):  
Systematic Review ◽  
Sensitivity And Specificity ◽  
Meta Analysis ◽  
False Negative ◽  
Significant Proportion ◽  
False Negative Rate ◽  
True Positive Rate ◽  
Axillary Staging ◽  
Positron Emission ◽  
Positive Rate

Background: This systematic review aimed at comparing performances of ultrasonography (US), magnetic resonance imaging (MRI), and fluorodeoxyglucose positron emission tomography (PET) for axillary staging, with a focus on micro- or micrometastases. Methods: A search for relevant studies published between January 2002 and March 2018 was conducted in MEDLINE database. Study quality was assessed using the QUality Assessment of Diagnostic Accuracy Studies checklist. Sensitivity and specificity were meta-analyzed using a bivariate random effects approach; Results: Across 62 studies (n = 10,374 patients), sensitivity and specificity to detect metastatic ALN were, respectively, 51% (95% CI: 43–59%) and 100% (95% CI: 99–100%) for US, 83% (95% CI: 72–91%) and 85% (95% CI: 72–92%) for MRI, and 49% (95% CI: 39–59%) and 94% (95% CI: 91–96%) for PET. Interestingly, US detects a significant proportion of macrometastases (false negative rate was 0.28 (0.22, 0.34) for more than 2 metastatic ALN and 0.96 (0.86, 0.99) for micrometastases). In contrast, PET tends to detect a significant proportion of micrometastases (true positive rate = 0.41 (0.29, 0.54)). Data are not available for MRI. Conclusions: In comparison with MRI and PET Fluorodeoxyglucose (FDG), US is an effective technique for axillary triage, especially to detect high metastatic burden without upstaging majority of micrometastases.

scholarly journals Macromolecule Particle Picking and Segmentation of a KLH Database by Unsupervised Cryo-EM Image Processing

Biomolecules ◽  
2019 ◽  
Vol 9 (12) ◽  
pp. 809
Author(s):  
Miguel Carrasco ◽  
Patricio Toledo ◽  
Nicole D. Tischler
Keyword(s):  
Signal To Noise Ratio ◽  
False Positive Rate ◽  
False Negative ◽  
False Negative Rate ◽  
Keyhole Limpet Hemocyanin ◽  
True Positive Rate ◽  
Cryo Electron Microscopy ◽  
Positive Rate ◽  
Top View ◽  
Non Negative Matrix Factorization

Segmentation is one of the most important stages in the 3D reconstruction of macromolecule structures in cryo-electron microscopy. Due to the variability of macromolecules and the low signal-to-noise ratio of the structures present, there is no generally satisfactory solution to this process. This work proposes a new unsupervised particle picking and segmentation algorithm based on the composition of two well-known image filters: Anisotropic (Perona–Malik) diffusion and non-negative matrix factorization. This study focused on keyhole limpet hemocyanin (KLH) macromolecules which offer both a top view and a side view. Our proposal was able to detect both types of views and separate them automatically. In our experiments, we used 30 images from the KLH dataset of 680 positive classified regions. The true positive rate was 95.1% for top views and 77.8% for side views. The false negative rate was 14.3%. Although the false positive rate was high at 21.8%, it can be lowered with a supervised classification technique.

scholarly journals KS(conf): A Light-Weight Test if a Multiclass Classifier Operates Outside of Its Specifications

2019 ◽  
Vol 128 (4) ◽  
pp. 970-995
Author(s):  
Rémy Sun ◽  
Christoph H. Lampert
Keyword(s):  
False Positive ◽  
False Positive Rate ◽  
A Priori ◽  
False Negative ◽  
True Positive Rate ◽  
Convolutional Network ◽  
The Road ◽  
Positive Rate ◽  
On The Road ◽  
Main Component

Abstract We study the problem of automatically detecting if a given multi-class classifier operates outside of its specifications (out-of-specs), i.e. on input data from a different distribution than what it was trained for. This is an important problem to solve on the road towards creating reliable computer vision systems for real-world applications, because the quality of a classifier’s predictions cannot be guaranteed if it operates out-of-specs. Previously proposed methods for out-of-specs detection make decisions on the level of single inputs. This, however, is insufficient to achieve low false positive rate and high false negative rates at the same time. In this work, we describe a new procedure named KS(conf), based on statistical reasoning. Its main component is a classical Kolmogorov–Smirnov test that is applied to the set of predicted confidence values for batches of samples. Working with batches instead of single samples allows increasing the true positive rate without negatively affecting the false positive rate, thereby overcoming a crucial limitation of single sample tests. We show by extensive experiments using a variety of convolutional network architectures and datasets that KS(conf) reliably detects out-of-specs situations even under conditions where other tests fail. It furthermore has a number of properties that make it an excellent candidate for practical deployment: it is easy to implement, adds almost no overhead to the system, works with any classifier that outputs confidence scores, and requires no a priori knowledge about how the data distribution could change.

scholarly journals A KNIME-Based Analysis of the Zebrafish Photomotor Response Clusters the Phenotypes of 14 Classes of Neuroactive Molecules

2015 ◽  
Vol 21 (5) ◽  
pp. 427-436 ◽  
Author(s):  
Daniëlle Copmans ◽  
Thorsten Meinl ◽  
Christian Dietz ◽  
Matthijs van Leeuwen ◽  
Julia Ortmann ◽  
...  
Keyword(s):  
Drug Discovery ◽  
False Negative ◽  
False Negative Rate ◽  
Automated Analysis ◽  
True Positive Rate ◽  
Zebrafish Embryos ◽  
True Positive ◽  
Neuroactive Compounds ◽  
Analysis Workflow ◽  
Positive Rate

Recently, the photomotor response (PMR) of zebrafish embryos was reported as a robust behavior that is useful for high-throughput neuroactive drug discovery and mechanism prediction. Given the complexity of the PMR, there is a need for rapid and easy analysis of the behavioral data. In this study, we developed an automated analysis workflow using the KNIME Analytics Platform and made it freely accessible. This workflow allows us to simultaneously calculate a behavioral fingerprint for all analyzed compounds and to further process the data. Furthermore, to further characterize the potential of PMR for mechanism prediction, we performed PMR analysis of 767 neuroactive compounds covering 14 different receptor classes using the KNIME workflow. We observed a true positive rate of 25% and a false negative rate of 75% in our screening conditions. Among the true positives, all receptor classes were represented, thereby confirming the utility of the PMR assay to identify a broad range of neuroactive molecules. By hierarchical clustering of the behavioral fingerprints, different phenotypical clusters were observed that suggest the utility of PMR for mechanism prediction for adrenergics, dopaminergics, serotonergics, metabotropic glutamatergics, opioids, and ion channel ligands.

Image Quality Assessment and Outliers Filtering in an Image-Based Animal Supervision System

Biometrics ◽  
2017 ◽  
pp. 1241-1257
Author(s):  
Ehsan Khoramshahi ◽  
Juha Hietaoja ◽  
Anna Valros ◽  
Jinhyeon Yun ◽  
Matti Pastell
Keyword(s):  
Image Quality ◽  
Quality Assessment ◽  
Real Time ◽  
Image Quality Assessment ◽  
False Negative ◽  
False Negative Rate ◽  
True Positive Rate ◽  
Supervision System ◽  
Positive Rate ◽  
Validation Set

This paper presents a probabilistic framework for the image quality assessment (QA), and filtering of outliers, in an image-based animal supervision system (asup). The proposed framework recognizes asup's imperfect frames in two stages. The first stage deals with the similarity analysis of the same-class distributions. The objective of this stage is to maximize the separability measures by defining a set of similarity indicators (SI) under the condition that the number of permissible values for them is restricted to be relatively low. The second stage, namely faulty frame recognition (FFR), deals with asup's QA training and real-time quality assessment (RTQS). In RTQS, decisions are made based on a real-time quality assessment mechanism such that the majority of the defected frames are removed from the consecutive sub routines that calculate the movements. The underlying approach consists of a set of SI indexes employed in a simple Bayesian inference model. The results confirm that a significant amount of defected frames can be efficiently classified by this approach. The performance of the proposed technique is demonstrated by the classification on a cross-validation set of mixed high and low quality frames. The classification shows a true positive rate of 88.6% while the false negative rate is only about 2.5%.

Image Quality Assessment and Outliers Filtering in an Image-Based Animal Supervision System

2018 ◽  
pp. 34-50
Author(s):  
Ehsan Khoramshahi ◽  
Juha Hietaoja ◽  
Anna Valros ◽  
Jinhyeon Yun ◽  
Matti Pastell
Keyword(s):  
Image Quality ◽  
Quality Assessment ◽  
Real Time ◽  
Image Quality Assessment ◽  
False Negative ◽  
False Negative Rate ◽  
True Positive Rate ◽  
Supervision System ◽  
Positive Rate ◽  
Validation Set

This paper presents a probabilistic framework for the image quality assessment (QA), and filtering of outliers, in an image-based animal supervision system (asup). The proposed framework recognizes asup's imperfect frames in two stages. The first stage deals with the similarity analysis of the same-class distributions. The objective of this stage is to maximize the separability measures by defining a set of similarity indicators (SI) under the condition that the number of permissible values for them is restricted to be relatively low. The second stage, namely faulty frame recognition (FFR), deals with asup's QA training and real-time quality assessment (RTQS). In RTQS, decisions are made based on a real-time quality assessment mechanism such that the majority of the defected frames are removed from the consecutive sub routines that calculate the movements. The underlying approach consists of a set of SI indexes employed in a simple Bayesian inference model. The results confirm that a significant amount of defected frames can be efficiently classified by this approach. The performance of the proposed technique is demonstrated by the classification on a cross-validation set of mixed high and low quality frames. The classification shows a true positive rate of 88.6% while the false negative rate is only about 2.5%.

scholarly journals Handling Imbalanced Class Problem of Measles InfectionRisk Prediction Model

2019 ◽  
Vol 9 (1) ◽  
pp. 3431-3435
Keyword(s):  
Logistic Regression ◽  
False Positive Rate ◽  
False Negative ◽  
False Negative Rate ◽  
Sampling Technique ◽  
True Positive Rate ◽  
True Positive ◽  
Positive Rate ◽  
Imbalanced Class ◽  
Medical Dataset

Measles is an emerging infectious disease with increasing number of reported cases. It is a vaccine-preventable disease;thus, it is common to have imbalanced class problem in the dataset. This study aims to resolve the imbalanced class problem for the prediction of measles infection risk and to compare the predictive results on a balanced dataset based on three machine learningtechniques. The data that was utilized in this study contained 37,884 records of suspected measles casesthat were highly imbalanced towards negative measles cases. The Synthetic Minority Over-Sampling Technique (SMOTE) was performed to balance thedistribution of the target attribute. The balanced dataset was then modelled using logistic regression, decision tree and Naïve Bayes. The predicted results indicated that logistic regression executed on the balanced dataset by SMOTE has the highest and most accurateclassification with 94.5% overall accuracy, 93.9% true positive rate, 5.8% false positive rate and 5.1% false negative rate. Therefore, SMOTE and other over-sampling approaches may be applicable to overcome imbalanced class issues in the medical dataset.

Image Quality Assessment and Outliers Filtering in an Image-Based Animal Supervision System

2015 ◽  
Vol 6 (2) ◽  
pp. 15-30
Author(s):  
Ehsan Khoramshahi ◽  
Juha Hietaoja ◽  
Anna Valros ◽  
Jinhyeon Yun ◽  
Matti Pastell
Keyword(s):  
Image Quality ◽  
Quality Assessment ◽  
Real Time ◽  
Image Quality Assessment ◽  
False Negative ◽  
False Negative Rate ◽  
True Positive Rate ◽  
Supervision System ◽  
Positive Rate ◽  
Validation Set

This paper presents a probabilistic framework for the image quality assessment (QA), and filtering of outliers, in an image-based animal supervision system (asup). The proposed framework recognizes asup's imperfect frames in two stages. The first stage deals with the similarity analysis of the same-class distributions. The objective of this stage is to maximize the separability measures by defining a set of similarity indicators (SI) under the condition that the number of permissible values for them is restricted to be relatively low. The second stage, namely faulty frame recognition (FFR), deals with asup's QA training and real-time quality assessment (RTQS). In RTQS, decisions are made based on a real-time quality assessment mechanism such that the majority of the defected frames are removed from the consecutive sub routines that calculate the movements. The underlying approach consists of a set of SI indexes employed in a simple Bayesian inference model. The results confirm that a significant amount of defected frames can be efficiently classified by this approach. The performance of the proposed technique is demonstrated by the classification on a cross-validation set of mixed high and low quality frames. The classification shows a true positive rate of 88.6% while the false negative rate is only about 2.5%.

scholarly journals LINKCALCULATOR – AN EFFICIENT LINK-BASED PHISHING DETECTION TOOL

2020 ◽  
Vol 4 (2) ◽  
pp. 37-44
Author(s):  
Orunsolu Abiodun ◽  
Sodiya A.S ◽  
Kareem S.O
Keyword(s):  
False Positive Rate ◽  
False Negative ◽  
False Negative Rate ◽  
True Positive Rate ◽  
True Negative ◽  
Negative Rate ◽  
Positive Rate ◽  
Link Information ◽  
Phishing Detection ◽  
Mobile Market

The problem of phishing attacks continues to demand new solutions as existing solutions are limited by various challenges such as high computational requirements, zero-day attacks, needs for updates, complex ruled-based, etc. Besides, the emerging mobile market demands simple solutions to phishing due to several factors such as memory, fragmentation, etc. In response to the above challenges, a simple anti-phishing tool called LinkCalculator is presented. The proposed LinkCalculator anti-phishing scheme is based on an algorithm designed to extract link characteristics from loading URLs to determine their legitimacy. Unlike the other link-based extraction approaches, the proposed approach introduced the concept of weight to represent the different links found in a URL. This is because certain link information within parsed webpages or requests is sufficient to classify them as phishing without loss of generality. The approach is experimented using a dataset of 300 instances consisting of 150 legitimate URLs and 150 phishing URLs from openly-available research datasets. The experimental results indicate a significance performance of 100%. True Negative Rate and 0.00% False Positive Rate for legitimate instances and True Positive Rate of 96.67% with 0.03 % False Negative Rate for phishing instances which indicate that the approach offers a more efficient lightweight approach to phishing detection.

Sign in / Sign up

Export Citation Format

Share Document