CE-BART: Cause-and-Effect BART for Visual Commonsense Generation.

"A Picture is worth a thousand words". Given an image, humans are able to deduce various cause-and-effect captions of past, current, and future events beyond the image. The task of visual commonsense generation has the aim of generating three cause-and-effect captions for a given image: (1) what needed to happen before, (2) what is the current intent, and (3) what will happen after. However, this task is challenging for machines, owing to two limitations: existing approaches (1) directly utilize conventional vision-language transformers to learn relationships between input modalities and (2) ignore relations among target cause-and-effect captions, but consider each caption independently. Herein, we propose Cause-and-Effect BART (CE-BART), which is based on (1) a structured graph reasoner that captures intra- and inter-modality relationships among visual and textual representations and (2) a cause-and-effect generator that generates cause-and-effect captions by considering the causal relations among inferences. We demonstrate the validity of CE-BART on the VisualCOMET and AVSD benchmarks. CE-BART achieved SOTA performance on both benchmarks, while an extensive ablation study and qualitative analysis demonstrated the performance gain and improved interpretability.

Dual-Scale Doppler Attention for Human Identification.

This paper considers a Deep Convolutional Neural Network (DCNN) with an attention mechanism referred to as Dual-Scale Doppler Attention (DSDA) for human identification given a micro-Doppler (MD) signature induced as input. The MD signature includes unique gait characteristics by different sized body parts moving, as arms and legs move rapidly, while the torso moves slowly. Each person is identified based on his/her unique gait characteristic in the MD signature. DSDA provides attention at different time-frequency resolutions to cater to different MD components composed of both fast-varying and steady. Through this, DSDA can capture the unique gait characteristic of each person used for human identification. We demonstrate the validity of DSDA on a recently published benchmark dataset, IDRad. The empirical results show that the proposed DSDA outperforms previous methods, using a qualitative analysis interpretability on MD signatures.

Effect of Cu Insertion into SnO2 Framework on Surface Properties and Carbonyl Sulfide Adsorption Performances.

The objective of this study is to introduce Cu into SnO2 sorbent for improving its COS adsorption capacity. Cu-doped SnO2 adsorbents were synthesized using a conventional sol-gel method with citric acid. X-ray diffraction studies revealed that up to 0.4 mol of Cu ions were well-inserted within the SnO2 framework. Scanning electron microscopy images confirmed that the addition of Cu ions reduced the particle size of the SnO2 sorbents. Additionally, it led to an increase in the Brunauer-Emmett-Teller surface area of the sorbents. The COS adsorption tests were carried out in the temperature range of 300-400 °C with a gas hourly space velocity of 8,500 h-1. It was found that Cu0.6SnO2 displayed higher COS adsorption capacity than the sorbents of other compositions, and the breakthrough time and COS adsorption capacity on it at 400 °;C were 170 min and 4.87 mg/g, respectively. X-ray photoelectron spectroscopy results indicated that the Cu2+ ions in the CuxSnO2 adsorbent converted into CuS by binding to the S2- ions in COS gas, while the remaining CO segments combined with the Sn atoms in SnO2 and then are adsorbed as SnCO. Overall, this study showed that the hard-soft acid-base rule is better followed in the Cu0.6SnO2 adsorbent than in the SnO2 adsorbent and that the adsorption is more stable.

Autologous Tissue Graft in Revision Medial Epicanthoplasty Using Subciliary Fat and Orbicularis Oculi Muscle.

The medial epicanthal fold is one of the racial anatomic characteristics of Asians. As medial epicanthoplasty has become one of the most common cosmetic surgeries among Asians, the need for revision of overcorrected medial epicanthus also increased. In revision medial epicanthoplasty, an autologous tissue graft to the subcutaneous plane is used to reduce postoperative scar. Medial epicanthoplasty, using V-Y advancement and rotation flap, was performed in 93 patients (revision medial epicanthoplasty with autologous tissue graft, 60 patients; revision medial epicanthoplasty only, 33 patients). A V-Y and rotation flap was designed to cover the overexposed lacrimal lake. A small amount of fat tissue and orbicularis oculi muscle were harvested from the upper eyelid through the double-fold line and grafted to the subcutaneous space of the medial epicanthal area. We evaluated the patients' satisfaction with the overall outcome and scar. The mean intercanthal distance increased from 32.3 mm before surgery to 34.6 mm after surgery. Satisfaction with the scar and overall outcome was higher in the patient group who underwent medial epicanthoplasty with autologous tissue graft. No complication was observed in the autologous tissue graft group, whereas 1 patient who underwent medial epicanthoplasty without autologous tissue graft showed recurrence of the medial epicanthal fold, 5 months after surgery. Medial epicanthoplasty with autologous tissue graft is a simple and reliable method to repair the overcorrected medial epicanthus that can efficiently increase the intercanthal distance and prevent scar contracture.

Secondary Cleft Nasal Deformity Correction Using Bioabsorbable Mesh.

INTRODUCTION: Septorhinoplasty in cleft patients is relatively difficult to perform and the outcomes are often unsatisfactory owing to weakness of the cartilage. Therefore, providing support to the cartilage is necessary. The authors used bioabsorbable mesh as a spreader or batten graft to support the septum in cleft patients those with weak cartilages. The purpose of this study was to evaluate the efficacy of bioabsorbable mesh when performing septorhinoplasty in cleft patients. METHODS: Between 2009 and 2013, 34 patients with cleft lip and nose underwent open septorhinoplasty by using bioabsorbable mesh at the Konkuk University Medical Center. The authors determined the nasolabial angle, Goode ratio, alar base-nasal tip-columellar base angle, and columellar axis angle and performed a panel evaluation. RESULTS: In all, 29 of the 34 patients were included; 5 patients were excluded as lost to follow-up. The nasolabial angle increased from 85.69° to 91.62° (P = 0.0032), and the mean Goode ratio increased from 0.51 to 0.55, which was approximately the ideal value (P = 0.0017). The alar base-nasal tip-columellar base angle decreased from 36.28 to 31.74 (P < 0.0001). The columellar axis angle increased from 82.36 to 86.90 (P < 0.0001). DISCUSSION: As bioabsorbable mesh can provide strong support, results in good esthetic outcomes, and causes minimal complications, it can be used in septorhinoplasty for cleft patients.

Scalable SoftGroup for 3D Instance Segmentation on Point Clouds.

This paper considers a network referred to as SoftGroup for accurate and scalable 3D instance segmentation. Existing state-of-the-art methods produce hard semantic predictions followed by grouping instance segmentation results. Unfortunately, errors stemming from hard decisions propagate into the grouping, resulting in poor overlap between predicted instances and ground truth and substantial false positives. To address the abovementioned problems, SoftGroup allows each point to be associated with multiple classes to mitigate the uncertainty stemming from semantic prediction. It also suppresses false positive instances by learning to categorize them as background. Regarding scalability, the existing fast methods require computational time on the order of tens of seconds on large-scale scenes, which is unsatisfactory and far from applicable for real-time. Our finding is that the k-Nearest Neighbor ( k-NN) module, which serves as the prerequisite of grouping, introduces a computational bottleneck. SoftGroup is extended to resolve this computational bottleneck, referred to as SoftGroup++. The proposed SoftGroup++ reduces time complexity with octree k-NN and reduces search space with class-aware pyramid scaling and late devoxelization. Experimental results on various indoor and outdoor datasets demonstrate the efficacy and generality of the proposed SoftGroup and SoftGroup++. Their performances surpass the best-performing baseline by a large margin (6%  âˆ¼  16%) in terms of AP 50. On datasets with large-scale scenes, SoftGroup++ achieves a 6× speed boost on average compared to SoftGroup. Furthermore, SoftGroup can be extended to perform object detection and panoptic segmentation with nontrivial improvements over existing methods.

Exposure-Response and Clinical Outcome Modeling of Inhaled Budesonide/Formoterol Combination in Asthma Patients.

Exposure-response and clinical outcome (CO) model for inhaled budesonide/formoterol was developed to quantify the relationship among pharmacokinetics (PK), pharmacodynamics (PD) and CO of the drugs and evaluate the covariate effect on model parameters. Sputum eosinophils cationic proteins (ECP) and forced expiratory volume (FEV1) were selected as PD markers and asthma control score was used as a clinical outcome. One- and two-compartment models were used to describe the PK of budesonide and formoterol, respectively. The indirect response model (IDR) was used to describe the PD effect for ECP and FEV1. In addition, the symptomatic effect on the disease progression model for CO was connected with IDR on each PD response. The slope for the effect of ECP and FEV1 to disease progression were estimated as 0.00008 and 0.644, respectively. Total five covariates (ex. ADRB2 genotype etc.) were searched using a stepwise covariate modeling method, however, there was no significant covariate effect. The results from the simulation study were showed that a 1 puff b.i.d. had a comparable effect of asthma control with a 2 puff b.i.d. As a result, the 1 puff b.i.d. of combination drug could be suggested as a standardized dose to minimize the side effects and obtain desired control of disease compared to the 2 puff b.i.d.

Flexible Piezoelectric Acoustic Sensors and Machine Learning for Speech Processing.

Flexible piezoelectric acoustic sensors have been developed to generate multiple sound signals with high sensitivity, shifting the paradigm of future voice technologies. Speech recognition based on advanced acoustic sensors and optimized machine learning software will play an innovative interface for artificial intelligence (AI) services. Collaboration and novel approaches between both smart sensors and speech algorithms should be attempted to realize a hyperconnected society, which can offer personalized services such as biometric authentication, AI secretaries, and home appliances. Here, representative developments in speech recognition are reviewed in terms of flexible piezoelectric materials, self-powered sensors, machine learning algorithms, and speaker recognition.

Electrochemical Synergies of Heterostructured Fe2O3-MnO Catalyst for Oxygen Evolution Reaction in Alkaline Water Splitting.

For efficient electrode development in an electrolysis system, Fe2O3, MnO, and heterojunction Fe2O3-MnO materials were synthesized via a simple sol-gel method. These particles were coated on a Ni-foam (NF) electrode, and the resulting material was used as an electrode to be used during an oxygen evolution reaction (OER). A 1000-cycle OER test in a KOH alkaline electrolyte indicated that the heterojunction Fe2O3-MnO/NF electrode exhibited the most stable and highest OER activity: it exhibited a low overvoltage (n) of 370 mV and a small Tafel slope of 66 mV/dec. X-ray photoelectron spectroscopy indicated that the excellent redox performance contributed to the synergy of Mn and Fe, which enhanced the OER performance of the Fe2O3-MnO/NF electrode. Furthermore, the effective redox reaction of Mn and Fe indicated that the structure maintained stability even under 1000 repeated OER cycles.

Application of Size and Maturation Functions to Population Pharmacokinetic Modeling of Pediatric Patients.

Traditionally, dosage for pediatric patients has been optimized using simple weight-scaled methods, but these methods do not always meet the requirements of children. To overcome this discrepancy, population pharmacokinetic (PK) modeling of size and maturation functions has been proposed. The main objective of the present study was to evaluate a new modeling method for pediatric patients using clinical data from three different clinical studies. To develop the PK models, a nonlinear mixed effect modeling method was employed, and to explore PK differences in pediatric patients, size with allometric and maturation with Michaelis-Menten type functions were evaluated. Goodness of fit plots, visual predictive check and bootstrap were used for model evaluation. Single application of size scaling to PK parameters was statistically significant for the over one year old group. On the other hand, simultaneous use of size and maturation functions was statistically significant for infants younger than one year old. In conclusion, population PK modeling for pediatric patients was successfully performed using clinical data. Size and maturation functions were applied according to established criteria, and single use of size function was applicable for over one year ages, while size and maturation functions were more effective for PK analysis of neonates and infants.