Perceptual metric-guided human image generation

Pose transfer, which synthesizes a new image of a target person in a novel pose, is valuable in several applications. Generative adversarial networks (GAN) based pose transfer is a new way for person re-identification (re-ID). Typical perceptual metrics, like Detection Score (DS) and Inception Score (IS), were employed to assess the visual quality after generation in pose transfer task. Thus, the existing GAN-based methods do not directly benefit from these metrics which are highly associated with human ratings. In this paper, a perceptual metrics guided GAN (PIGGAN) framework is proposed to intrinsically optimize generation processing for pose transfer task. Specifically, a novel and general model-Evaluator that matches well the GAN is designed. Accordingly, a new Sort Loss (SL) is constructed to optimize the perceptual quality. Morevover, PIGGAN is highly flexible and extensible and can incorporate both differentiable and indifferentiable indexes to optimize the attitude migration process. Extensive experiments show that PIGGAN can generate photo-realistic results and quantitatively outperforms state-of-the-art (SOTA) methods.

Predicting Alzheimer's disease from Clinical Dementia Rating, Estimated total intracranial volume & Atlas Scaling Factor : Incorporating Ensemble approach into Automated Machine Learning

The current state-of-the-art for automated machine learning is adopted to predict Alzheimer's disease (AD) by adopting variables such as Mini Mental State Examination score, estimated total intracranial volume and Atlas Scaling Factor. A macro-weighted average Area under the Response-operating Curve of 0.96 is achieved with a close-to-perfect AD detection score after incorporating the ensemble approach. Such predictive models shall serve to optimize risk stratification and management protocols for this enfeebling ailment.

Research on the Fastest Detection Method for Weak Trends under Noise Interference

Trend anomaly detection is the practice of comparing and analyzing current and historical data trends to detect real-time abnormalities in online industrial data-streams. It has the advantages of tracking a concept drift automatically and predicting trend changes in the shortest time, making it important both for algorithmic research and industry. However, industrial data streams contain considerable noise that interferes with detecting weak anomalies. In this paper, the fastest detection algorithm “sliding nesting” is adopted. It is based on calculating the data weight in each window by applying variable weights, while maintaining the method of trend-effective integration accumulation. The new algorithm changes the traditional calculation method of the trend anomaly detection score, which calculates the score in a short window. This algorithm, SNWFD–DS, can detect weak trend abnormalities in the presence of noise interference. Compared with other methods, it has significant advantages. An on-site oil drilling data test shows that this method can significantly reduce delays compared with other methods and can improve the detection accuracy of weak trend anomalies under noise interference.

Meta-analytic Re-assessment of the Validity of Miller Forensic Assessment Test for Detection of Malingering

Background: The Miller Forensic Assessment of Symptoms Test (M-FAST) is used widely for the assessment of malingering of medical symptoms. Its validity has allegedly been supported by meta-analytic study of M-FAST in 2019 by Detullio et al. Credibility of Detullio’s results is damaged by an inclusion of data based on analog validation and also on dubious convergent validation procedures that falsify estimates of M-FAST’s validity. Method: In the present study, the meta-analysis was calculated on 3 types of M-FAST data: (1) 5 samples of scores of healthy persons instructed to respond honestly, (2) 5 samples of scores of medical patients, and (3) 10 samples of scores of healthy persons instructed to feign mental illness (so called “instructed malingerers”). Results: In an ANOVA (F(2,815)=398.50, p<.0001), significantly lowest M-FAST scores were those of healthy controls (mean=1.59, SD=2.80), the next significantly higher scores were those of legitimate patients (mean=4.85, SD=4.22), and the instructed malingerers had significantly highest scores (mean=12.34, SD=5.71). Discussion: The significant difference between healthy controls and patients shows that inferences from analog validations of the M-FAST are inherently false. Furthermore, data of legitimate patients with severe psychiatric illness suggested that they may face the risk of about 50% to be falsely classified as malingerers by the M-FAST. Moreover, almost all validations of the M-FAST were done only with “instructed malingerers” (healthy volunteers instructed to feign symptoms). This overestimates the test’s capacity to detect real-life malingerers. Montes and Guyton documented that “instructed malingerers” warned to avoid detection score much lower than the unwarned ones (effect size: Cohen’s d=3.05). M-FAST’s capacity for detection of real-life malingerers may be extremely low, in particular those more genuinely motivated to evade detection, well prepared, better educated, and systematically feigning only a few specific symptoms such as depression, pain, and insomnia. Conclusion: The M-FAST should no longer be used.

Multi-Sector Oriented Object Detector for Accurate Localization in Optical Remote Sensing Images

Oriented object detection in optical remote sensing images (ORSIs) is a challenging task since the targets in ORSIs are displayed in an arbitrarily oriented manner and on small scales, and are densely packed. Current state-of-the-art oriented object detection models used in ORSIs primarily evolved from anchor-based and direct regression-based detection paradigms. Nevertheless, they still encounter a design difficulty from handcrafted anchor definitions and learning complexities in direct localization regression. To tackle these issues, in this paper, we proposed a novel multi-sector oriented object detection framework called MSO2-Det, which quantizes the scales and orientation prediction of targets in ORSIs via an anchor-free classification-to-regression approach. Specifically, we first represented the arbitrarily oriented bounding box as four scale offsets and angles in four quadrant sectors of the corresponding Cartesian coordinate system. Then, we divided the scales and angle space into multiple discrete sectors and obtained more accurate localization information by a coarse-granularity classification to fine-grained regression strategy. In addition, to decrease the angular-sector classification loss and accelerate the network’s convergence, we designed a smooth angular-sector label (SASL) that smoothly distributes label values with a definite tolerance radius. Finally, we proposed a localization-aided detection score (LADS) to better represent the confidence of a detected box by combining the category-classification score and the sector-selection score. The proposed MSO2-Det achieves state-of-the-art results on three widely used benchmarks, including the DOTA, HRSC2016, and UCAS-AOD data sets.

Development and Validation of a Simplified Probability Assessment Score Integrated With Age‐Adjusted d ‐Dimer for Diagnosis of Acute Aortic Syndromes

Background When acute aortic syndromes (AASs) are suspected, pretest clinical probability assessment and d ‐dimer (DD) testing are diagnostic options allowing standardized care. Guidelines suggest use of a 12‐item/3‐category score (aortic dissection detection) and a DD cutoff of 500 ng/mL. However, a simplified assessment tool and a more specific DD cutoff could be advantageous. Methods and Results In a prospective derivation cohort (n=1848), 6 items identified by logistic regression (thoracic aortic aneurysm, severe pain, sudden pain, pulse deficit, neurologic deficit, hypotension), composed a simplified score (AORTAs) assigning 2 points to hypotension and 1 to the other items. AORTAs≤1 and ≥2 defined low and high clinical probability, respectively. Age‐adjusted DD was calculated as years/age × 10 ng/mL (minimum 500). The AORTAs score and AORTAs≤1/age‐adjusted DD rule were validated in 2 patient cohorts: a high‐prevalence retrospective cohort (n=1035; 22% AASs) and a low‐prevalence prospective cohort (n=447; 11% AASs) subjected to 30‐day follow‐up. The AUC of the AORTAs score was 0.729 versus 0.697 of the aortic dissection detection score ( P =0.005). AORTAs score assessment reclassified 16.6% to 25.1% of patients, with significant net reclassification improvement of 10.3% to 32.7% for AASs and −8.6 to −17% for alternative diagnoses. In both cohorts, AORTAs≥2 had superior sensitivity and slightly lower specificity than aortic dissection detection ≥2. In the prospective validation cohort, AORTAs≤1/age‐adjusted DD had a sensitivity of 100%, a specificity of 48.6%, and an efficiency of 43.3%. Conclusions AORTAs is a simplified score with increased sensitivity, improved AAS classification, and minor trade‐off in specificity, amenable to integration with age‐adjusted DD for diagnostic rule‐out.

Anomaly Event Detection in Security Surveillance Using Two-Stream Based Model

Anomaly event detection has been extensively researched in computer vision in recent years. Most conventional anomaly event detection methods can only leverage the single-modal cues and not deal with the complementary information underlying other modalities in videos. To address this issue, in this work, we propose a novel two-stream convolutional networks model for anomaly detection in surveillance videos. Specifically, the proposed model consists of RGB and Flow two-stream networks, in which the final anomaly event detection score is the fusion of those of two networks. Furthermore, we consider two fusion situations, including the fusion of two streams with the same or different number of layers respectively. The design insight is to leverage the information underlying each stream and the complementary cues of RGB and Flow two-stream sufficiently. Two datasets (UCF-Crime and ShanghaiTech) are used to validate the effectiveness of proposed solution.

On the Propagation of Satellite Precipitation Estimation Errors: From Passive Microwave to Infrared Estimates

ABSTRACTThe launch of NOAA’s latest generation of geostationary satellites known as the Geostationary Operational Environmental Satellite (GOES)-R Series has opened new opportunities in quantifying precipitation rates. Recent efforts have strived to utilize these data to improve space-based precipitation retrievals. The overall objective of the present work is to carry out a detailed error budget analysis of the improved Self-Calibrating Multivariate Precipitation Retrieval (SCaMPR) algorithm for GOES-R and the passive microwave (MW) combined (MWCOMB) precipitation dataset used to calibrate it with an aim to provide insights regarding strengths and weaknesses of these products. This study systematically analyzes the errors across different climate regions and also as a function of different precipitation types over the conterminous United States. The reference precipitation dataset is Ground-Validation Multi-Radar Multi-Sensor (GV-MRMS). Overall, MWCOMB reveals smaller errors as compared to SCaMPR. However, the analysis indicated that that the major portion of error in SCaMPR is propagated from the MWCOMB calibration data. The major challenge starts with poor detection from MWCOMB, which propagates in SCaMPR. In particular, MWCOMB misses 90% of cool stratiform precipitation and the overall detection score is around 40%. The ability of the algorithms to quantify precipitation amounts for the Warm Stratiform, Cool Stratiform, and Tropical/Stratiform Mix categories is poor compared to the Convective and Tropical/Convective Mix categories with additional challenges in complex terrain regions. Further analysis showed strong similarities in systematic and random error models with both products. This suggests that the potential of high-resolution GOES-R observations remains underutilized in SCaMPR due to the errors from the calibrator MWCOMB.