Deep Feature Statistics Mapping for Generalized Screen Content Image Quality Assessment.

The statistical regularities of natural images, referred to as natural scene statistics, play an important role in no-reference image quality assessment. However, it has been widely acknowledged that screen content images (SCIs), which are typically computer generated, do not hold such statistics. Here we make the first attempt to learn the statistics of SCIs, based upon which the quality of SCIs can be effectively determined. The underlying mechanism of the proposed approach is based upon the mild assumption that the SCIs, which are not physically acquired, still obey certain statistics that could be understood in a learning fashion. We empirically show that the statistics deviation could be effectively leveraged in quality assessment, and the proposed method is superior when evaluated in different settings. Extensive experimental results demonstrate the Deep Feature Statistics based SCI Quality Assessment (DFSS-IQA) model delivers promising performance compared with existing NR-IQA models and shows a high generalization capability in the cross-dataset settings. The implementation of our method is publicly available at https://github.com/Baoliang93/DFSS-IQA.

Exceedances of Secondary Aerosol Formation from In-Use Natural Gas Heavy-Duty Vehicles Compared to Diesel Heavy-Duty Vehicles.

This work, for the first time, assessed the secondary aerosol formation from both in-use diesel and natural gas heavy-duty vehicles of different vocations when they were operated on a chassis dynamometer while the vehicles were exercised on different driving cycles. Testing was performed on natural gas vehicles equipped with three-way catalysts (TWCs) and diesel trucks equipped with diesel oxidation catalysts, diesel particulate filters, and selective catalytic reduction systems. Secondary aerosol was measured after introducing dilute exhaust into a 30 m3 environmental chamber. Particulate matter ranged from 0.18 to 0.53 mg/mile for the diesel vehicles vs 1.4-85 mg/mile for the natural gas vehicles, total particle number ranged from 4.01 × 1012 to 3.61 × 1013 for the diesel vehicles vs 5.68 × 1012-2.75 × 1015 for the natural gas vehicles, and nonmethane organic gas emissions ranged from 0.032 to 0.05 mg/mile for the diesel vehicles vs 0.012-1.35 mg/mile for the natural gas vehicles. Ammonia formation was favored in the TWC and was found in higher concentrations for the natural gas vehicles (ranged from ∼0 to 1.75 g/mile) than diesel vehicles (ranged from ∼0 to 0.4 g/mile), leading to substantial secondary ammonium nitrate formation (ranging from 8.5 to 98.8 mg/mile for the natural gas vehicles). For the diesel vehicles, one had a secondary ammonium nitrate of 18.5 mg/mile, while the other showed essentially no secondary ammonium nitrate formation. The advanced aftertreatment controls in diesel vehicles resulted in almost negligible secondary organic aerosol (SOA) formation (ranging from 0.046 to 2.04 mg/mile), while the natural gas vehicles led to elevated SOA formation that was likely sourced from the engine lubricating oil (ranging from 3.11 to 39.7 mg/mile). For two natural gas vehicles, the contribution of lightly oxidized lubricating oil in the primary organic aerosol was dominant (as shown in the mass spectra analysis), leading to enhanced SOA mass. Heavily oxidized lubricating oil was also observed to contribute to the SOA formation for other natural gas vehicles.

No-Reference Image Quality Assessment by Hallucinating Pristine Features.

In this paper, we propose a no-reference (NR) image quality assessment (IQA) method via feature level pseudo-reference (PR) hallucination. The proposed quality assessment framework is rooted in the view that the perceptually meaningful features could be well exploited to characterize the visual quality, and the natural image statistical behaviors are exploited in an effort to deliver the accurate predictions. Herein, the PR features from the distorted images are learned by a mutual learning scheme with the pristine reference as the supervision, and the discriminative characteristics of PR features are further ensured with the triplet constraints. Given a distorted image for quality inference, the feature level disentanglement is performed with an invertible neural layer for final quality prediction, leading to the PR and the corresponding distortion features for comparison. The effectiveness of our proposed method is demonstrated on four popular IQA databases, and superior performance on cross-database evaluation also reveals the high generalization capability of our method. The implementation of our method is publicly available on https://github.com/Baoliang93/FPR.

The impact of hydrogenated vegetable oil (HVO) on the formation of secondary organic aerosol (SOA) from in-use heavy-duty diesel vehicles.

This manuscript contains an assessment of tailpipe emissions and secondary aerosol formation from two in-use heavy-duty diesel vehicles (HDDVs) with different aftertreatment systems when operated with ultra-low sulfur diesel (ULSD) and hydrogenated vegetable oil (HVO) operated on a chassis dynamometer. Secondary aerosol formation was characterized from the HDDVs' diluted exhaust collected and photochemically aged in a 30 m3 mobile atmospheric chamber. Primary nitrogen oxide (NOx) and particulate matter (PM) emissions were reduced for both vehicles operating on HVO compared to ULSD. For the vehicles with no selective catalytic reduction (SCR) system, secondary aerosol production was ~2 times higher for ULSD compared to HVO. The composition of primary aerosol was exclusively organic for the vehicle with no SCR system regardless of fuel type. The composition of secondary aerosol with HVO was primarily organic for the vehicle equipped with diesel particulate filter (DPF)/SCR system; however, when the same vehicle was tested with ULSD, the composition was ~20% organic (80% ammonium nitrate). The results reported here revealed that the in-use vehicle with no-SCR had a non-functioning DPF leading to dramatic increases in secondary aerosol formation when compared to the DPF/SCR vehicle. The high-resolution mass spectra analysis showed that the POA of HVO combustion contained relatively lower portion of CH class compounds (or higher CHO class compounds) compared to ULSD under the similar conditions, which can be rationalized by the higher cetane number of HVO. Substantial growth of oxidized organic aerosol (such as m/z 44 peak) were observed after 5 h of photochemical oxidation, consistent with aged organic aerosols present in the atmosphere. The C4H9+ fragment at m/z 57 peak was used as a tracer to calculate evolution of secondary organic aerosol formation.

Real-world NOx emissions from heavy-duty diesel, natural gas, and diesel hybrid electric vehicles of different vocations on California roadways.

This study assessed the real-world nitrogen oxide (NOx) emissions from 50 heavy-duty vehicles of different vocations and engine technologies using portable emissions measurement systems (PEMS). This is one of the most comprehensive in-use emissions studies conducted to date, which played a key role in the development of CARB's (California Air Recourses Board) updated EMission FACtor (EMFAC) model, especially for natural gas vehicles. In-use emissions testing was performed on school and transit buses, refuse haulers, goods movement vehicles, and delivery vehicles while were driven over their normal operating routes in the South Coast Air Basin. Engine technologies included diesel engines with and without selective catalytic reduction (SCR) systems, compressed natural gas (CNG) engines and liquified petroleum gas (LPG) engines, and SCR-equipped diesel hybrid electric vehicles. For most vehicles, the in-use NOx emissions were higher than the certification standards for the engine. Diesel vehicles generally showed higher brake-specific NOx emissions compared to the CNG vehicles. NOx emissions were strongly dependent on the SCR temperature, with SCR temperatures below 200 °C resulting in elevate brake-specific NOx. The 0.02 g/bhp-hr certified CNG vehicles showed the largest reductions in NOx emissions. The diesel hybrid electric vehicles showed important distance-specific NOx benefits compared to the conventional diesel vehicles, but higher emissions compared to the CNG and LPG vehicles. Overall, average NOx reductions were 75%, 94%, 65%, 79%, respectively, for the 0.2 CNG, 0.02 CNG, diesel hybrid electric, and LPG vehicles compared to diesel vehicles, due in part to some diesel vehicles with particularly high emissions, indicating that the widespread implementation of advanced technology and alternative fuel vehicles could provide important NOx reductions and a path for meeting air quality targets in California and elsewhere.

Superpixel-Based Quality Assessment of Multi-Exposure Image Fusion for Both Static and Dynamic Scenes.

Multi-exposure image fusion (MEF) algorithms have been used to merge a stack of low dynamic range images with various exposure levels into a well-perceived image. However, little work has been dedicated to predicting the visual quality of fused images. In this work, we propose a novel and efficient objective image quality assessment (IQA) model for MEF images of both static and dynamic scenes based on superpixels and an information theory adaptive pooling strategy. First, with the help of superpixels, we divide fused images into large- and small-changed regions using the structural inconsistency map between each exposure and fused images. Then, we compute the quality maps based on the Laplacian pyramid for large- and small-changed regions separately. Finally, an information theory induced adaptive pooling strategy is proposed to compute the perceptual quality of the fused image. Experimental results on three public databases of MEF images demonstrate the proposed model achieves promising performance and yields a relatively low computational complexity. Additionally, we also demonstrate the potential application for parameter tuning of MEF algorithms.

On-road gaseous and particulate emissions from GDI vehicles with and without gasoline particulate filters (GPFs) using portable emissions measurement systems (PEMS).

This study assessed the on-road gaseous and particulate emissions from three current technology gasoline direct injection (GDI) vehicles using portable emissions measurement systems (PEMS). Two vehicles were also retrofitted with catalyzed gasoline particulate filters (GPFs). All vehicles were exercised over four routes with different topological and environmental characteristics, representing urban, rural, highway, and high-altitude driving conditions. The results showed strong reductions in particulate mass (PM), soot mass, and particle number emissions with the use of GPFs. Particle emissions were found to be highest during urban and high-altitude driving compared to highway driving. The reduction efficiency of the GPFs ranged from 44% to 99% for overall soot mass emissions. Similar efficiencies were found for particle number and PM mass emissions. In most cases, nitrogen oxide (NOx) emissions showed improvements with the catalyzed GPFs in the underfloor position with the additional catalytic volume. No significant differences were seen in carbon dioxide (CO2) and carbon monoxide (CO) emissions with the vehicles retrofitted with GPFs.

Deep Guided Learning for Fast Multi-Exposure Image Fusion.

We propose a fast multi-exposure image fusion (MEF) method, namely MEF-Net, for static image sequences of arbitrary spatial resolution and exposure number. We first feed a low-resolution version of the input sequence to a fully convolutional network for weight map prediction. We then jointly upsample the weight maps using a guided filter. The final image is computed by a weighted fusion. Unlike conventional MEF methods, MEF-Net is trained end-to-end by optimizing the perceptually calibrated MEF structural similarity (MEF-SSIM) index over a database of training sequences at full resolution. Across an independent set of test sequences, we find that the optimized MEF-Net achieves consistent improvement in visual quality for most sequences, and runs 10 to 1000 times faster than state-of-the-art methods. The code is made publicly available at.

Perceptual Evaluation for Multi-Exposure Image Fusion of Dynamic Scenes.

A common approach to high dynamic range (HDR) imaging is to capture multiple images of different exposures followed by multi-exposure image fusion (MEF) in either radiance or intensity domain. A predominant problem of this approach is the introduction of the ghosting artifacts in dynamic scenes with camera and object motion. While many MEF methods (often referred to as deghosting algorithms) have been proposed for reduced ghosting artifacts and improved visual quality, little work has been dedicated to perceptual evaluation of their deghosting results. Here we first construct a database that contains 20 multiexposure sequences of dynamic scenes and their corresponding fused images by nine MEF algorithms. We then carry out a subjective experiment to evaluate fused image quality, and find that none of existing objective quality models for MEF provides accurate quality predictions. Motivated by this, we develop an objective quality model for MEF of dynamic scenes. Specifically, we divide the test image into static and dynamic regions, measure structural similarity between the image and the corresponding sequence in the two regions separately, and combine quality measurements of the two regions into an overall quality score. Experimental results show that the proposed method significantly outperforms the state-of-the-art. In addition, we demonstrate the promise of the proposed model in parameter tuning of MEF methods.1.

[Analysis of drug-induced acute renal failure in Shanghai].

OBJECTIVE: To investigate the incidence and prognosis of drug-induced acute renal failure (ARF) in Shanghai. METHODS: The registration forms of ARF patients admitted in 17 hospitals of and over the middle class in Shanghai from January 1, 2004 to December 31, 2006 were screened prospectively. The data, such as epidemiology, survival, mortality, and morbidity were analyzed. RESULTS: 347 of the 1200 ARF patients (28.9%), 224 males and 123 females, aged (58+/-20), suffered from drug-induced ARF. 51.0% of the 347 patients were older than 60. 60.2% of the drug-induced ARF in the non-surgical departments were community-acquired, while 55.7% of the drug-induced ARF in the surgical departments were hospital-acquired. Among the non-surgical departments, the incidence of hospital-acquired drug-induced ARF was the lowest in the department of nephrology (9.5%), while higher in the departments of hematology, cardiology, and neurology, and among the surgical departments, it was the lowest in department of renal surgery, while higher in the departments of liver transplantation, neurosurgery, and cardiovascular surgery. The most common complication was chronic kidney disease (CKD) (n=69, 19.9%), followed by cerebrovascular disease (n=59, 17.0%), diabetes mellitus (n=43, 12.4%), and hypertension (n=41, 11.8%). Renal biopsy showed acute tubular necrosis (18, 37.5%), acute interstitial nephritis (11, 22.9%), and acute infectious tubulo-interstitial nephritis (6, 12.5%). Antibiotics (47.8%) were the head causes of drug-induced ARF, especially aminoglycoside (17.0%) and cephalosporins (12.7%), followed by diuretics (22.2%) and radiocontrasts (13.3%). 22.5% of the drug-induced ARF patients had used two or more drugs. 119 patients (34.3%) needed renal replacement treatment. 100 of the 347 patients (28.8%) died. 188 of the surviving patients (54.2%) had their renal function recovered completely, the renal function of 42 of them (12.1%) was recovered partially, and 17 of then (4.9%) required dialysis when discharged. CONCLUSION: Drug-induced ARF is common with higher incidence in the patients with complications. Antibiotics, diuretic agents, and contrast medium are the main causes.