Explore the vast range of titles available.

Deep models have been studied in point cloud classification for the applications of autonomous driving and robotics. One challenging issue is that the point cloud of the same object could be discrepantly captured depending on sensors. Such a difference is the main cause of the domain gap. The deep models trained with one domain of point clouds may not work well with other domains because of such a domain gap. A technique to reduce domain inconsistency is domain adaptation. In this paper, we propose an unsupervised domain adaptation with two novel schemes. First, to improve unreliable pseudo-label assignment, we introduce a voting-based procedure based on the recycling max pooling module, which involves self-paced learning. It helps to increase the training stability of the models. Second, to learn the geometrical characteristics of point clouds in unfamiliar settings, we propose a training method of cutting plane identification, which works in an unsupervised way. Testing with the popular point cloud dataset of PointDA-10 and Sim-to-Real, experiments show that our method increase classification accuracy by 6.5%-points on average, ModelNet and ShapeNet as the source domain and ScanNet, and ScanObjectNN as the target domain. From an ablation study, it was observed that each method contributes to improving the robustness of domain adaptation.

Background: In bottom-up proteomics using data-independent acquisition mass spectrometry (DIA-MS), quantitative measurements are obtained following multiple steps of protein fragmentation and ionization, which introduces cumulative errors and impairs the effectiveness of classical statistical methods. This study proposes an alternative statistical approach for testing group mean differences at the peptide level in quantitative bottom-up proteomics. Results: We present a novel probabilistic graphical model, that accounts for the non-normality of empirical distributions and the correlations between fragment ion quantities. Based on the model, we propose a new statistical method that improves upon the classical feature-based approach by incorporating distribution-free shrinkage estimation of covariance matrices and bootstrap-based estimation of degrees-of-freedom. Simulated experiments demonstrate that the proposed method outperforms the four most widely used classical methods in terms of specificity, sensitivity, and accuracy, particularly when the data distribution closely resembles real MS data, and under conditions of small sample sizes. Numerical analysis of real quantitative tandem mass spectrometry data reveals that the proposed method effectively identifies candidate peptides exhibiting changes in mean quantity following treatment with the kinase inhibitor Staurosporine. Conclusions: The proposed statistical method offers an effective alternative to classical approaches for differential analysis of peptides in quantitative bottom-up proteomics using DIA-MS. The R software package MDstatsDIAMS is available at https://github.com/namgillee/MDstatsDIAMS .

Skewness and obliqueness of vehicle plate images influence license plate recognition. The more tilted plate images are, the harder the recognition task is. To this end, if plate images are preprocessed to be aligned and rectified, the recognition performance would improve. We propose deep neural network models that can locate four corner plate positions, which can then be used to perform the perspective transformation that can be used to rectify plates. Such a transformation is called homography. The models consist of two sequential parts: a feature extraction part having convolution and a regression part with fully connected layers. The models are open in the sense that the feature extraction part can host other well-known models such as Mobilenet as long as they have the feature capture capability. We devise a loss function as the sum of Euclidean distance between predicted coordinates and ground truth and discuss image augmentation schemes. The experiment results show that the models with well-known object detection models are able to predict corner positions with relatively high precision.

Biomedical databases grow by more than a thousand new publications every day. The large volume of biomedical literature that is being published at an unprecedented rate hinders the discovery of relevant knowledge from keywords of interest to gather new insights and form hypotheses. A text-mining tool, PubTator, helps to automatically annotate bioentities, such as species, chemicals, genes, and diseases, from PubMed abstracts and full-text articles. However, the manual re-organization and analysis of bioentities is a non-trivial and highly time-consuming task. ChexMix was designed to extract the unique identifiers of bioentities from query results. Herein, ChexMix was used to construct a taxonomic tree with allied species among Korean native plants and to extract the medical subject headings unique identifier of the bioentities, which co-occurred with the keywords in the same literature. ChexMix discovered the allied species related to a keyword of interest and experimentally proved its usefulness for multi-species analysis.

The pH-selective interaction between the immunoglobulin G (IgG) fragment crystallizable region (Fc region) and the neonatal Fc receptor (FcRn) is critical for prolonging the circulating half-lives of IgG molecules through intracellular trafficking and recycling. By using directed evolution, we successfully identified Fc mutations that improve the pH-dependent binding of human FcRn and prolong the serum persistence of a model IgG antibody and an Fc-fusion protein. Strikingly, trastuzumab-PFc29 and aflibercept-PFc29, a model therapeutic IgG antibody and an Fc-fusion protein, respectively, when combined with our engineered Fc (Q311R/M428L), both exhibited significantly higher serum half-lives in human FcRn transgenic mice than their counterparts with wild-type Fc. Moreover, in a cynomolgus monkey model, trastuzumab-PFc29 displayed a superior pharmacokinetic profile to that of both trastuzumab-YTE and trastuzumab-LS, which contain the well-validated serum half-life extension Fcs YTE (M252Y/S254T/T256E) and LS (M428L/N434S), respectively. Furthermore, the introduction of two identified mutations of PFc29 (Q311R/M428L) into the model antibodies enhanced both complement-dependent cytotoxicity and antibody-dependent cell-mediated cytotoxicity activity, which are triggered by the association between IgG Fc and Fc binding ligands and are critical for clearing cancer cells. In addition, the effector functions could be turned off by combining the two mutations of PFc29 with effector function-silencing mutations, but the antibodies maintained their excellent pH-dependent human FcRn binding profile. We expect our Fc variants to be an excellent tool for enhancing the pharmacokinetic profiles and potencies of various therapeutic antibodies and Fc-fusion proteins. Drug development: antibodies engineered with superior drug properties The precise addition of two mutations in the “tail” region of therapeutic antibodies could lead to more potent and long-lived drugs. Professor Sang Taek Jung (Korea University, South Korea) and co-workers generated antibodies with mutations in the “Fc” region, a part of the antibody that interacts with cell surface receptors to mediate immune activation. The researchers identified one particular pair of mutations that improved binding to a receptor involved in protecting antibodies from degradation. Drugs built around this doubly mutated Fc had prolonged half-lives in transgenic mice and monkeys. The mutations also enhanced the ability of antibody drugs to eliminate target cells—a property that, if therapeutically appropriate, could be silenced via additional antibody engineering without affecting the drugs’ extended half-life.

Network-based methods for the analysis of drug-target interactions have gained attention and rely on the paradigm that a single drug can act on multiple targets rather than a single target. In this study, we have presented a novel approach to analyze the interactions between the chemicals in the medicinal plants and multiple targets based on the complex multipartite network of the medicinal plants, multi-chemicals, and multiple targets. The multipartite network was constructed via the conjunction of two relationships: chemicals in plants and the biological actions of those chemicals on the targets. In doing so, we introduced an index of the efficacy of chemicals in a plant on a protein target of interest, called target potency score (TPS). We showed that the analysis can identify specific chemical profiles from each group of plants, which can then be employed for discovering new alternative therapeutic agents. Furthermore, specific clusters of plants and chemicals acting on specific targets were retrieved using TPS that suggested potential drug candidates with high probability of clinical success. We expect that this approach may open a way to predict the biological functions of multi-chemicals and multi-plants on the targets of interest and enable repositioning of the plants and chemicals.

In this study, a flexible conductive filament was fabricated by mixing thermoplastic polyurethane (TPU), carbon nanotubes (CNTs), and Cu powder. A polymer extruder was used to disperse conductive materials in the TPU matrix. Because the dispersion of CNTs in polymers is difficult, the dispersion process was repeated several times for the homogeneity of the fabricated conductive filament. Cu powder with high electrical conductivity was additionally added to improve the electrical characteristics of the conductive polymer. As TPU generally has excellent ductility and durability, the fabricated filament can secure flexibility. The resistance was measured and compared according to the mixing ratio of CNT/Cu powder. Based on the obtained results, the resistance of the conductive filament decreased as the CNT and Cu powder contents were increased. The addition of metal powders, such as CNTs and Cu powder, however, reduced the ductility of TPU. Finally, the developed conductive filament was used to fabricate a simple closed photodiode circuit.

The experiments of wire arc additive manufacturing (WAAM) with general pulsed gas metal arc welding (GMAW) machine were performed to optimize process parameters. As optimized conditions, 140A welding current, 34 V welding voltage, 40 cm/min welding speed, and 15 mm contact tip to work distance (CTWD), were selected. Welding current and voltage affected heat input. Lower deposition, short circuit mode, and spatters were obtained by insufficient heat input. The melting of previous layer and lower height were obtained by excessive heat input. Welding speed and CTWD affected surface state and arc stability. Unsound surface and bead shape due to short circuit mode and spatters were observed in slow welding speed and short CTWD conditions. Arc instability were also observed in fast welding speed and long CTWD conditions. With the optimized process parameter, curved propeller shape blade was manufactured by the change of robot arm teaching point. It can be considered that the desired sufficient quality was obtained due to curved shape.

Several studies have been conducted to predict issues caused by the welding process. The finite element method is extremely useful in analysing this welding process. Generally, in the analysis of welded structures, the welded joint is assumed to be rigid to large structure on vehicle-body crash or fatigue. This assumption does not require a theoretical background of a complicated welding process; therefore, it is relatively easy to set up an analysis model, which significantly reduces the analysis time. However, material properties and mechanical behavior of welded structures cannot be expressed, and there is no sufficient verification for this in several industries. Moreover, few studies have reported on strength analysis, including the welding process; therefore, this study presents a simple modeling using commercial program based on the finite element method. The welding process is simulated by identifying the welding-heat-source variable through experiments and a value close to the measured deposited metal and fusion-zone size are obtained. A flow stress–strain curve based on a uniaxial tensile test is obtained and the tensile strength of the welded specimen is analyzed.

The orientation of plate images in license plate recognition is one of the factors that influence its accuracy. In particular, tilted plate images are harder to detect and recognize characters with than aligned ones. To this end, the rectification of plates in a preprocessing step is essential to improve their performance. We propose deep models to estimate four-corner coordinates of tilted plates. Since the predicted corners can then be used to rectify plate images, they can help improve plate recognition in plate recognition. The main contributions of this work are a set of open-structured hybrid networks to predict corner positions and a novel loss function that combines pixel-wise differences with position-wise errors, producing performance improvements. Regarding experiments using proprietary plate images, one of the proposed modes produces a 3.1% improvement over the established warping method.