Retrieval Setting Research Articles

B - 125 Atypical Alzheimer’s Disease in a Highly Educated Individual: a Case Study

Abstract Objective Early onset of atypical Alzheimer’s disease (ad) can be challenging to identify, especially in highly educated individuals. Protein deposits in neural areas other than the medial temporal lobe can be associated with a heterogeneous neurocognitive pattern. We present a case that highlights the value of biomarker studies in these individuals. Method A 67-year-old, right-handed female with 18 years of education presented with inefficient word retrieval and mental set loss, which emerged approximately two to three years prior. Symptoms worsened in the context of gabapentin use status post lung cancer resection without return to baseline with steady progression. Remaining medical history includes a 1.4 cm right parietal falcine meningioma, hyperlipidemia, and depression. Results In the context of high average premorbid functioning, challenges included complex executive functioning, subtle visuospatial deficits, and encoding and recall inefficiencies. Naming was low average. Neuroimaging revealed mild parietal and dorsal frontal atrophy, potentially age-appropriate, with no significant hippocampal atrophy. Conclusions The present case highlights the possibility of atypical protein deposits and neurocognitive pattern in a highly educated individual with the potential of an early onset ad process. The value of biomarkers and neuropsychological nuances will be of increasing relevance in the context of emerging amyloid therapies.

Updated OMI Glyoxal Column Measurements Using Collection 4 Level 1B Radiances

AbstractThis study presents new glyoxal (CHOCHO) products from the Ozone Monitoring Instrument (OMI) by utilizing updated level 1B irradiance/radiance data (Collection 4) and an updated glyoxal retrieval algorithm. The adoption of Collection 4 contributes to the reduction of artificial signals in differential glyoxal slant column densities (dSCDs) and improved fitting root mean square, and the updated retrieval settings result in fewer negative values of glyoxal dSCDs over oceans and less noisy dSCDs in the South Atlantic Anomaly. On‐line calculations of air mass factors consider interactive physical processes between input parameters. To address persistent trends in glyoxal SCDs over the Pacific Ocean that remain despite these updates, a trend correction is implemented. We evaluate the updated OMI glyoxal products using inter‐comparisons with GOME‐2A/2B glyoxal products. OMI glyoxal products exhibit good spatial and temporal agreement with GOME‐2A/2B, with correlation coefficients of 0.75–0.78 globally and 0.84–0.85 over source regions. Small biases are observed in OMI glyoxal vertical column densities, ranging from −0.2 ± 5.7% to 9 ± 3% in low and high glyoxal conditions, respectively, against GOME‐2A/2B. These advancements contribute to the reliability and accuracy of OMI glyoxal products, enhancing their utility for atmospheric studies and enabling a 20‐year‐long data record suitable for climate studies.

The Far-INfrarEd Spectrometer for Surface Emissivity (FINESSE) – Part 2: First measurements of the emissivity of water in the far-infrared

Abstract. In this paper, we describe a method for retrieving the surface emissivity of specular surfaces across the wavenumber range of 400–1600 cm−1 using novel radiance measurements of the Far-INfrarEd Spectrometer for Surface Emissivity (FINESSE) instrument. FINESSE is described in detail in Part 1 (Murray et al., 2024) of this paper. We apply the method to two sets of measurements of distilled water. The first set of emissivity retrievals is of distilled water heated above ambient temperature to enhance the signal-to-noise ratio. The second set of emissivity retrievals is of ambient temperate water at a range of viewing angles. In both cases, the observations agree well with calculations based on compiled refractive indices across the mid- and far-infrared. It is found that the reduced contrast between the up- and downwelling radiation in the ambient temperature case degrades the performance of the retrieval. Therefore, a filter is developed to target regions of high contrast, which improves the agreement between the ambient temperature emissivity retrieval and the predicted emissivity. These retrievals are, to the best of our knowledge, the first published simultaneous retrievals of the surface temperature and emissivity of water that extend into the far-infrared and demonstrate a method that can be used and further developed for the in situ retrieval of the emissivity of other surfaces in the field.

Efficacy of refrigerated gel packs for therapeutic hypothermia in neonatal retrieval: a retrospective cohort study

ObjectiveTo determine the efficacy of refrigerated gel packs in achieving and maintaining target temperature in neonates receiving therapeutic hypothermia (TH) for hypoxic ischaemic encephalopathy during neonatal retrieval.DesignRetrospective cohort study.SettingPaediatric Infant...

Ventilator-assisted preoxygenation in an aeromedical retrieval setting.

Ventilator-assisted preoxygenation (VAPOX) is a method of preoxygenation and apnoeic ventilation which has been tried in hospital setting. We aimed to describe VAPOX during intubation of critically unwell patients in aeromedical retrieval setting. Retrospective observational study of VAPOX performed at LifeFlight Retrieval Medicine (LRM) between January 2018 and December 2022 across Queensland, Australia. Demographic and clinical data were recorded. Descriptive statistics and paired Student's t-tests were used to evaluate the efficacy of VAPOX on oxygen saturation (SpO2). VAPOX was used in 40 patients. Diagnoses included pneumonia (n = 11), COPD (n = 6) and neurological (n = 7). Patients were intubated in hospital (n = 36), in helicopter (n = 2) and ambulance (n = 2). Median VAPOX settings were: positive end-expiratory pressure 6 (IQR 5-9), pressure support 10 (IQR 10-14) and back up respiratory rate 14 (IQR 11-18). Twelve agitated patients underwent delayed sequence induction with ketamine. There was a statistically significant increase in SpO2 after application of VAPOX (P < 0.001), followed by a slight decrease after intubation (P = 0.006). Mean SpO2 were significantly improved after intubation compared with on arrival of LRM (P = 0.016). Hypotension was present prior to VAPOX (n = 13), during VAPOX (n = 2) and post-intubation (n = 15). Two patients had cardiac arrest. Three patients were started on VAPOX but subsequently failed. There were no significant oxygen depletion or aspiration events. VAPOX can be considered for pre-intubation optimisation in the retrieval environment. The incidence of post-intubation critical hypoxia was low, and hypotension was high. Pre-intubation respiratory physiology can be optimised by delivering variable pressure supported minute ventilation, achieving a low incidence of critical hypoxia.

Early experience in use of videolaryngoscopy by a neonatal pre-hospital and retrieval service.

To describe initial experience with use of the Glidescope Go videolaryngoscope by an Australian neonatal pre-hospital and retrieval service. We conducted a 31-month retrospective review of an airway registry for neonates intubated by MedSTAR Kids clinicians. Twenty-two patients were intubated using the Glidescope Go, compared with 50 using direct laryngoscopy. First-pass success was 17/22 (77.3%) with the Glidescope Go and 38/50 (76%) with direct laryngoscopy. Complications occurred in 7/22 (32%) and 8/50 (16%), respectively. On initial review of this practice change, videolaryngoscopy allows neonatal tracheal intubation with a comparable success rate to direct laryngoscopy in a pre-hospital and retrieval setting.

EESSO: Exploiting Extreme and Smooth Signals via Omni-frequency learning for Text-based Person Retrieval

Most of the existing methods manage to tackle the problem of text-based person retrieval from the spatial-wise perspective. In this paper, we manage to address the problem of this task from a novel perspective, namely, the frequency-wise perspective. To this end, we propose to Exploit Extreme and Smooth Signals via Omni-frequency learning (EESSO) through a jointly optimized multi-stream architecture. It consists of a Spatial Information Stream (SIS), an Extreme Signal Stream (ESS) and a Smooth Signal Stream (SSS). EESSO aims to excavate the complementary effect between spatial-wise and frequency-wise features, so as to achieve a superior performance. A novel Uncertainty-Guided Mutual Learning Mechanism (UG-MLM) is utilized during training, which not only enables the three streams to communicate with and learn from each other, but also models the data-related heteroscedastic uncertainty as a weight for knowledge transference, and hence enables each stream to adaptively allocate knowledge from the other two streams. A large number of experiments are carried out on the widely-used CUHK-PEDES, RSTPReid and ICFG-PEDES datasets to verify the effectiveness of EESSO. Through the experimental results, it may not be hard to find that EESSO has achieved the state-of-the-art performance in supervised, weakly supervised and cross-domain text-based person retrieval settings.

DOMINER: Domain Feature Mining from Unstructured Data for Effective Text Summarization

In the current era of internet, a huge volume of opinionated sentiment reviews is generated on a daily basis. Normally, such reviews are enriched with deep sentiments and express the opinions on various features of a product or service under consideration. Identification of a set of important domain features opens an opportunity to build automatic product and service summarization system. It is observed that the common phenomena is to consider the frequently appearing words as the domainb features. However, not all frequently appearing words can be domain-specific features. In this paper, a novel Domain Feature Miner (DOMINER) approach is proposed for robust extractive summarization. The entire domain feature set mining problem is modelled as a clustering problem. Firstly, the bond energy based clustering technique is employed to cluster the domain features based on their frequency and co-appearance counts. Later, relevant clusters are extracted for the final set of domain feature set retrieval. The Proposed DOMINER scheme is extensively evaluated against quantitative performance metrics such as precision, recall, and F-score for six diversified domains such as Cellphone, Camera, Laptop, Tablet, Television, and Hotel. Experimental results on benchmark data sets reveal that proposed DOMINER scheme mines high quality of domain feature from unstructured reviews evident from precision, recall, and F-score values as 78.10%, 64.21%, and 70.48%, respectively against state of the art existing schemes. The high quality domain features extraction from DOMINER help generate the robust extractive summaries for product and services.

A new accurate retrieval algorithm of bromine monoxide columns inside minor volcanic plumes from Sentinel-5P TROPOMI observations

Abstract. Bromine monoxide (BrO) is a key radical in the atmosphere, influencing the chemical state of the atmosphere, most notably the abundance of ozone (O3). O3 depletion caused by the release of bromine has been observed and modeled in polar regions, salt pans, and in particular inside volcanic plumes. Furthermore, the molar ratio of BrO and SO2 – which can be detected simultaneously via spectroscopic measurements using the differential optical absorption spectroscopy (DOAS) method – is a proxy for the magmatic composition of a volcano and potentially an eruption forecast parameter. The detection of BrO in volcanic plumes from satellite spectroscopic observations is limited by the precision and sensitivity of the retrieval, which so far only allowed for the detection of BrO during major eruptions. The unprecedented spatial resolution of up to 3.5 km×5.5 km and the high signal-to-noise ratio of the TROPOspheric Monitoring Instrument (TROPOMI) on board Sentinel-5 Precursor (S-5P) enable observing and monitoring volcanic bromine release globally even for minor eruptions or even quiescent degassing. In this study, we investigate how far the BrO retrieval can be improved using TROPOMI data and how well BrO can be detected, even in small eruptions and during quiescent volcanic degassing. There are two steps for which improvements in accuracy are investigated and applied: the improvement and quantitative determination of (1) the detection limit of the DOAS BrO column retrieval and (2) the correction of the non-volcanic background BrO signal. First, the DOAS retrieval settings are varied, and their influence on accuracy and precision is investigated with respect to the detection limit and potential systematic influences. Based on these results, we propose a dedicated DOAS evaluation scheme optimized for the detection of BrO in volcanic plumes. For the DOAS retrieval, we propose the use of a large fit window from 323–360 nm, yielding a statistical uncertainty lower by a factor of 1.8 compared to previous BrO DOAS algorithms while not enhancing systematic influences. Second, the effect of the background BrO is reduced by a latitude-dependent empirical correction scheme correlated to cloud information as well as information on the O3 column. Via these improvements, the combined statistical and systematic uncertainties in the resulting BrO vertical column density is on the order of 7×1012moleculescm-2. We present a new and accurate retrieval algorithm of BrO columns from TROPOMI observations which allows for the detection of even slightly enhanced BrO amounts inside minor eruptive plumes of bromine-rich volcanoes. While designed specifically for TROPOMI observations, the retrieval algorithm is in general also applicable to other hyperspectral satellite observations. However, some parts might require adaptation.

Reasoning over Public and Private Data in Retrieval-Based Systems

Abstract Users an organizations are generating ever-increasing amounts of private data from a wide range of sources. Incorporating private context is important to personalize open-domain tasks such as question-answering, fact-checking, and personal assistants. State-of-the-art systems for these tasks explicitly retrieve information that is relevant to an input question from a background corpus before producing an answer. While today’s retrieval systems assume relevant corpora are fully (e.g., publicly) accessible, users are often unable or unwilling to expose their private data to entities hosting public data. We define the Split Iterative Retrieval (SPIRAL) problem involving iterative retrieval over multiple privacy scopes. We introduce a foundational benchmark with which to study SPIRAL, as no existing benchmark includes data from a private distribution. Our dataset, ConcurrentQA, includes data from distinct public and private distributions and is the first textual QA benchmark requiring concurrent retrieval over multiple distributions. Finally, we show that existing retrieval approaches face significant performance degradations when applied to our proposed retrieval setting and investigate approaches with which these tradeoffs can be mitigated. We release the new benchmark and code to reproduce the results.1

Technical note: Constraining the hydroxyl (OH) radical in the tropics with satellite observations of its drivers – first steps toward assessing the feasibility of a global observation strategy

Abstract. Despite its importance in controlling the abundance of methane (CH4) and a myriad of other tropospheric species, the hydroxyl radical (OH) is poorly constrained due to its large spatial heterogeneity and the inability to measure tropospheric OH with satellites. Here, we present a methodology to infer tropospheric column OH (TCOH) in the tropics over the open oceans using a combination of a machine learning model, output from a simulation of the GEOS model, and satellite observations. Our overall goals are to assess the feasibility of our methodology, to identify potential limitations, and to suggest areas of improvement in the current observational network. The methodology reproduces the variability of TCOH from independent 3D model output and of observations from the Atmospheric Tomography mission (ATom). While the methodology also reproduces the magnitude of the 3D model validation set, the accuracy of the magnitude when applied to observations is uncertain because current observations are insufficient to fully evaluate the machine learning model. Despite large uncertainties in some of the satellite retrievals necessary to infer OH, particularly for NO2 and formaldehyde (HCHO), current satellite observations are of sufficient quality to apply the machine learning methodology, resulting in an error comparable to that of in situ OH observations. Finally, the methodology is not limited to a specific suite of satellite retrievals. Comparison of TCOH determined from two sets of retrievals does show, however, that systematic biases in NO2, resulting both from retrieval algorithm and instrumental differences, lead to relative biases in the calculated TCOH. Further evaluation of NO2 retrievals in the remote atmosphere is needed to determine their accuracy. With slight modifications, a similar methodology could likely be expanded to the extratropics and over land, with the benefits of increasing our understanding of the atmospheric oxidation capacity and, for instance, informing understanding of recent CH4 trends.

Orthonormal product quantization network for scalable face image retrieval

Existing deep quantization methods provided an efficient solution for large-scale image retrieval. However, the significant intra-class variations, like pose, illumination, and expressions in face images, still pose a challenge. In light of this, face image retrieval requires sufficiently powerful learning metrics, which are absent in current deep quantization works. Moreover, to tackle the growing unseen identities in the query stage, face image retrieval drives more demands regarding model generalization and scalability than general image retrieval tasks. This paper integrates product quantization with orthonormal constraints into an end-to-end deep learning framework to effectively retrieve face images. Specifically, we propose a novel scheme that uses predefined orthonormal vectors as codewords to enhance the quantization informativeness and reduce codewords’ redundancy. A tailored loss function maximizes discriminability among identities in each quantization subspace for both the quantized and original features. An entropy-based regularization term is imposed to reduce the quantization error. Experiments are conducted on four commonly-used face datasets under both seen and unseen identity retrieval settings. Our method outperforms all the compared state-of-the-art under both settings. The proposed orthonormal codewords consistently boost both models’ standard retrieval performance and generalization ability, demonstrating the superiority of our method for scalable face image retrieval.

A meta-analysis of effects and safety of Tripterygium wilfordii polyglycoside in the treatment of IgA nephropathy.

To evaluate the effects and safety of Tripterygium wilfordii polyglycoside (TWP) in the treatment of immunoglobulin A (IgA) nephropathy. A computer-assisted study search of Chinese Biomedical Database (CBM), Chinese Journal Full-text Database (CNKI), Wanfang Database, Chinese Scientific Journal Database (VIP), PubMed, Medline, EMBASE and Cochrane Library was performed, with the time range of retrieval set between the establishment of the database to December 31, 2019. Articles of randomized controlled trials on the treatment of IgA nephropathy by Tripterygium wilfordii polyglycoside were collected, and then screened according to the inclusion and exclusion criteria. Next, the quality of the papers was assessed, effective data were extracted, and a meta-analysis of the included studies was conducted using the Review Manager 5.3 software provided by the Cochrane Collaboration. Thirty randomized controlled trials (RCT) were included ultimately, and the meta-analysis showed that 1) Single (Sgl) TWP group was superior to angiotensin-converting enzyme inhibitor/angiotension receptor blocker (ACEI/ARB) group in terms of complete remission [odds ratio (OR) = 4.74, p-value < 0.00001], total remission (OR = 3.90, p-value < 0.0001), 24-hour proteinuria [mean difference (MD) = 1.18, p-value < 0.00001], and serum albumin (MD = - 8.23, p-value < 0.00001), and no significant difference in serum creatinine (MD = 2.09, p-value = 0.08) was found between Sgl TWP and control groups; TWP + ACEI/ARB group was superior in complete remission (OR = 2.57, p-value < 0.00001), total remission (OR = 4.36, p-value < 0.00001), serum albumin [standardized mean difference (SMD) = -0.68, p-value = 0.0005], 24-hour proteinuria (SMD = 1.24, p-value < 0.00001) and serum creatinine (SMD = 0.48, p-value = 0.006); 2) TWP group was superior to glucocorticoid group in complete remission (OR = 1.93, p-value < 0.0010), total remission (OR = 3.71, p-value < 0.00001), serum albumin (MD = -3.50, p-value = 0.002), 24-hour proteinuria (SMD = 0.93, p-value < 0.0001) and serum creatinine (SMD = 0.88, p-value = 0.006); 3) TWP group was better than mycophenolate mofetil (MMF) group in complete remission (OR = 2.05, p = 0.005), total remission (OR = 3.30, p-value = 0.002), 24-hour proteinuria (MD = 2.61, p-value < 0.0001), and serum albumin (MD = -6.43, p-value < 0.00001), but the differences in serum creatinine (MD = 1.28, p-value = 0.89) between TWP and control groups were not significant. Besides, TWP + ACEI/ARB group had a higher adverse reaction rate than the control group (OR = 2.21, p-value = 0.04), but there was no significant difference in the adverse reaction rate between other control and experimental groups (p-value > 0.05). The present evidence shows that Tripterygium wilfordii polyglycoside can effectively improve the remission rate, reduce proteinuria, and protect kidney function of IgA nephropathy patients, and also has good safety. However, limited by the quality of the included studies, the effects and safety of Tripterygium wilfordii polyglycoside in the treatment of IgA nephropathy need to be verified by more high-quality, large-scale, multi-center RCTs.

A peer-to-peer file storage and sharing system based on consortium blockchain

In the era of big data, data is playing an increasingly important role in scientific study, and reliable storage and secure sharing of data have become a research hotspot. At present, centralized solutions based on data centers and cloud storage have problems with data-right confirmation and center trust. A large number of decentralized storage solutions are public systems, in which blockchain technology, as a tool for value exchange, does not solve the problems of data verification and system supervision. We propose a peer-to-peer storage system with identity access, which achieves data validation, cross-organizational data retrieval, trusted authorization, and sharing based on the consortium blockchain. Our solution proposes a peer-to-peer data storage scheme based on the consortium blockchain and a set of identity authentication mechanisms compatible with the consortium blockchain. Based on this, we propose a blockchain-based permission control scheme and a set of retrieval, authorization, and sharing processes. Finally, we implemented and tested the system to prove the feasibility of the scheme.

Passive ground-based remote sensing of radiation fog

Abstract. Accurate boundary layer temperature and humidity profiles are crucial for successful forecasting of fog, and accurate retrievals of liquid water path are important for understanding the climatological significance of fog. Passive ground-based remote sensing systems such as microwave radiometers (MWRs) and infrared spectrometers like the Atmospheric Emitted Radiance Interferometer (AERI), which measures spectrally resolved infrared radiation (3.3 to 19.2 µm), can retrieve both thermodynamic profiles and liquid water path. Both instruments are capable of long-term unattended operation and have the potential to support operational forecasting. Here we compare physical retrievals of boundary layer thermodynamic profiles and liquid water path during 12 cases of thin (LWP&lt;40 g m−2) supercooled radiation fog from an MWR and an AERI collocated in central Greenland. We compare both sets of retrievals to in-situ measurements from radiosondes and surface-based temperature and humidity sensors. The retrievals based on AERI observations accurately capture shallow surface-based temperature inversions (0–10 m a.g.l.) with lapse rates of up to −1.2 ∘C m−1, whereas the strength of the surface-based temperature inversions retrieved from MWR observations alone are uncorrelated with in-situ measurements, highlighting the importance of constraining MWR thermodynamic profile retrievals with accurate surface meteorological data. The retrievals based on AERI observations detect fog onset (defined by a threshold in liquid water path) earlier than those based on MWR observations by 25 to 185 min. We propose that, due to the high sensitivity of the AERI instrument to near-surface temperature and small changes in liquid water path, the AERI (or an equivalent infrared spectrometer) could be a useful instrument for improving fog monitoring and nowcasting, particularly for cases of thin radiation fog under otherwise clear skies, which can have important radiative impacts at the surface.

SHREC’22 track: Open-Set 3D Object Retrieval

This paper reports the results of the SHREC’22 track: Open-Set 3D Object Retrieval, the goal of which is to evaluate the performance of different retrieval algorithms under the Open-Set setting and modality-missing setting, respectively. Since objects from unseen categories are very common in real-world applications, we design the open-set 3D object retrieval to expand the application of traditional 3D object retrieval. In this track, we generate open-set 3D object retrieval datasets OS-MN40 and OS-MN40-Miss based on the ModelNet40 dataset, which are collected for the open-set setting and both open-set setting and modality-missing setting, respectively. Both the two datasets include the training set (2822 objects from 8 categories) and the retrieval set (960 query objects and 8527 target objects from the other 32 categories). The categories of retrieval (query/target) sets are not seen in the training set. For each object in the OS-MN40, four types of modalities, including mesh, point cloud, multi-view, and voxel, are provided. Each object in the OS-MN40-Miss is represented with incomplete modality information, which is collected to simulate the retrieval task in the real world. This track attracted eight participants from four countries and 191 runs of all submissions. The evaluation results show a promising scenario about open-set retrieval on 3D objects with multi-modal and multi-resolution representation and reveal interesting insights in dealing with retrieving 3D objects in unknown-category objects.

Quantifying the impact of meteorological uncertainty on emission estimates and the risk to aviation using source inversion for the Raikoke 2019 eruption

Abstract. Due to the remote location of many volcanoes, there is substantial uncertainty about the timing, amount and vertical distribution of volcanic ash released when they erupt. One approach to determine these properties is to combine prior estimates with satellite retrievals and simulations from atmospheric dispersion models to create posterior emission estimates, constrained by both the observations and the prior estimates, using a technique known as source inversion. However, the results are dependent not only on the accuracy of the prior assumptions, the atmospheric dispersion model and the observations used, but also on the accuracy of the meteorological data used in the dispersion simulations. In this study, we advance the source inversion approach by using an ensemble of meteorological data from the Met Office Global and Regional Ensemble Prediction System to represent the uncertainty in the meteorological data and apply it to the 2019 eruption of Raikoke. Retrievals from the Himawari-8 satellite are combined with NAME dispersion model simulations to create posterior emission estimates. The use of ensemble meteorology provides confidence in the posterior emission estimates and associated dispersion simulations that are used to produce ash forecasts. Prior mean estimates of fine volcanic ash emissions for the Raikoke eruption based on plume height observations are more than 15 times higher than any of the mean posterior ensemble estimates. In addition, the posterior estimates have a different vertical distribution, with 27 %–44 % of ash being emitted into the stratosphere compared to 8 % in the mean prior estimate. This has consequences for the long-range transport of ash, as deposition to the surface from this region of the atmosphere happens over long timescales. The posterior ensemble spread represents uncertainty in the inversion estimate of the ash emissions. For the first 48 h following the eruption, the prior ash column loadings lie outside an estimate of the error associated with a set of independent satellite retrievals, whereas the posterior ensemble column loadings do not. Applying a risk-based methodology to an ensemble of dispersion simulations using the posterior emissions shows that the area deemed to be of the highest risk to aviation, based on the fraction of ensemble members exceeding predefined ash concentration thresholds, is reduced by 49 %. This is compared to estimates using an ensemble of dispersion simulations using the prior emissions with ensemble meteorology. If source inversion had been used following the eruption of Raikoke, it would have had the potential to significantly reduce disruptions to aviation operations. The posterior inversion emission estimates are also sensitive to uncertainty in other eruption source parameters and internal dispersion model parameters. Extending the ensemble inversion methodology to account for uncertainty in these parameters would give a more complete picture of the emission uncertainty, further increasing confidence in these estimates.

Dynamic Motion Matching: Design and Implementation of a Context-Aware Animation System for Games

Despite modern game systems adopting motion matching to retrieve an appropriate short motion clip from a database in real-time, existing methods struggle to support complex gaming scenes due to their inability to adapt live the motion retrieval based on the context. This paper presents the design and implementation of a context-aware character animation system, synthesizing realistic animations suitable for complex game scenes from a large-scale motion database. This system, called dynamic motion matching (DyMM), enables geometry and objects aware motion synthesis by introducing a two-phase context computation: an offline subspace decomposition of motion clips for creating a set of retrieval sub-spaces tailored to specific contexts and a subspace ensemble matching to compare relevant sub-features to determine the most appropriate motion clip. We also show the system architecture and implementation details applicable to a production-grade game engine. We verified the effectiveness of our method with industry-level motion data captured by professional game artists for multiple configurations and character controllers. The results of this study show that, by finding motion clips that comply well with the scene context, one can leverage large motion capture datasets to create practical systems that generate believable and controllable animations for games.

Effects of COVID-19 pandemic restrictions on an Australian neonatal and paediatric retrieval service.

AimThe COVID‐19 pandemic and associated travel and social distancing restrictions have reduced paediatric intensive care unit admissions for respiratory illnesses. The effects on retrieval (transport) services remain unquantified. Our study examined the utility of statistical process control in assessing the impact of the COVID‐19 pandemic on the number of neonatal and paediatric transfers in an Australian retrieval service.MethodsData collected prospectively from the SA Ambulance Service MedSTAR Emergency Retrieval database in South Australia were analysed from January 2015 to June 2021. Statistical process control methodology, a combination of a time series analysis and assessment for common and special cause variation, was used to assess the impact of the COVID‐19 pandemic on retrieval workload (primary outcome of interest).ResultsA total of 5659 neonatal and paediatric transfers occurred during the study period and were included. A significant decrease in paediatric transfers occurred after the initial lockdown measures in March 2020 were announced in South Australia (special cause variation). However, a similar reduction was not observed for neonatal transfers (common cause variation).ConclusionOur study demonstrates that statistical process control may be effectively used to understand the effects of external events and processes on usual activity patterns in the retrieval setting. We found a reduction in retrieval numbers for paediatric transfers but no effect on neonatal transfer numbers. The decline in paediatric transfers was primarily attributed to reduced respiratory cases.

Local Semantic Correlation Modeling Over Graph Neural Networks for Deep Feature Embedding and Image Retrieval.

Deep feature embedding aims to learn discriminative features or feature embeddings for image samples which can minimize their intra-class distance while maximizing their inter-class distance. Recent state-of-the-art methods have been focusing on learning deep neural networks with carefully designed loss functions. In this work, we propose to explore a new approach to deep feature embedding. We learn a graph neural network to characterize and predict the local correlation structure of images in the feature space. Based on this correlation structure, neighboring images collaborate with each other to generate and refine their embedded features based on local linear combination. Graph edges learn a correlation prediction network to predict the correlation scores between neighboring images. Graph nodes learn a feature embedding network to generate the embedded feature for a given image based on a weighted summation of neighboring image features with the correlation scores as weights. Our extensive experimental results under the image retrieval settings demonstrate that our proposed method outperforms the state-of-the-art methods by a large margin, especially for top-1 recalls.