One fundamental problem in Earth Vision is to accurately find the locations and identify the categories of the interesting objects in the aerial images, for which oriented bounding boxes (OBBs) are ...usually employed to depict better the objects emerging with arbitrary orientations. However, the regression of the OBBs always suffers from the ambiguous problem in the definition of the regression targets, which often reduces the convergency efficiency and decreases the detection accuracy. Although there are some methods like the binary segmentation map that can handle this problem, it brings a new problem of ambiguous background pixels in the OBBs. In this article, we propose to cast the OBB regression as a center-probability-map (CenterMap)-prediction problem, thus largely eliminating the ambiguities on the target definitions and the background pixels. The predicted CenterMaps are then used to generate the OBBs. The CenterMap OBB representation is simple, yet effective. Furthermore, to distinguish better the interesting objects from the cluttered background, a weighted pseudosegmentation-guided attention network is adopted to provide the object-level features for predicting the horizontal bounding boxes and the OBBs. The experimental results on three widely used data sets, i.e., DOTA, HRSC2016, and UCAS-AOD, demonstrate the effectiveness of our proposed method.
The recent advances in sequencing technologies enable the assembly of individual genomes to the quality of the reference genome. How to integrate multiple genomes from the same species and make the ...integrated representation accessible to biologists remains an open challenge. Here, we propose a graph-based data model and associated formats to represent multiple genomes while preserving the coordinate of the linear reference genome. We implement our ideas in the minigraph toolkit and demonstrate that we can efficiently construct a pangenome graph and compactly encode tens of thousands of structural variants missing from the current reference genome.
Along with the concept of circular economy growing worldwide, circular business models (CBMs) have been receiving ever greater attention in both the business sector and academia. However, the ...existing literature on the CBM is scattered and fragmented; this study offers an integrated firm‐level framework to link CBM typologies, the circular economy transition process, and relevant tools for CBM development and clarifies the positioning and roles of those tools in the process. In response to the fragmentation issue, results of this study are presented in three subtopics: (a) CBM typologies and archetypes, (b) transition guidelines, and (c) major analytical tools for CBM research. The roles and functions of CBM typologies and tools were integrated in different stages of the transition process, and the challenges and shortfalls for CBM research in the various stages were identified. This work lays the foundation for future operational studies.
Despite numerous ion beam irradiation of cell experiments performed over the past five decades, the relationship between the biological effectiveness of ion beams and the physical characteristics of ...the ion beam remains unclear. Using 1,118 sets of
in vitro
cell survival experiments with ion beam irradiation, compiled by the Particle Irradiation Data Ensemble (PIDE) project, the relationship between cell survival and the fluence and linear energy transfer (LET) of the ion beam was established. Unlike previous studies, the closed-form analytical function is independent of photon irradiation and takes a universal form across all ion and cell species. A new understanding of the biological effectiveness of ion beams is crucial for predicting tumor response and toxicities in ion beam radiation therapy, along with radiation protection for high-LET ion beams with low fluence.
•TgNN model trained with data while being guided by theory of the underlying problem.•TgNN achieves better predictability, reliability, and generalizability than DNN.•TgNN tested for cases with ...changed BCs, noisy data or outliers, and engineering controls.
Active researches are currently being performed to incorporate the wealth of scientific knowledge into data-driven approaches (e.g., neural networks) in order to improve the latter’s effectiveness. In this study, the Theory-guided Neural Network (TgNN) is proposed for deep learning of subsurface flow. In the TgNN, as supervised learning, the neural network is trained with available observations or simulation data while being simultaneously guided by theory (e.g., governing equations, other physical constraints, engineering controls, and expert knowledge) of the underlying problem. The TgNN can achieve higher accuracy than the ordinary Deep Neural Network (DNN) because the former provides physically feasible predictions and can be more readily generalized beyond the regimes covered with the training data. Furthermore, the TgNN model is proposed for subsurface flow with heterogeneous model parameters. Several numerical cases of two-dimensional transient saturated flow are introduced to test the performance of the TgNN. In the learning process, the loss function contains data mismatch, as well as PDE constraint, engineering control, and expert knowledge. After obtaining the parameters of the neural network by minimizing the loss function, a TgNN model is built that not only fits the data, but also adheres to physical/engineering constraints. Predicting the future response can be easily realized by the TgNN model. In addition, the TgNN model is tested in more complicated scenarios, such as prediction with changed boundary conditions, learning from noisy data or outliers, transfer learning, learning from sparse data, and engineering controls. Numerical results demonstrate that the TgNN model achieves much better predictability, reliability, and generalizability than DNN models due to the physical/engineering constraints in the former.
In recent years, deep learning has presented a great advance in the hyperspectral image (HSI) classification. Particularly, long short-term memory (LSTM), as a special deep learning structure, has ...shown great ability in modeling long-term dependencies in the time dimension of video or the spectral dimension of HSIs. However, the loss of spatial information makes it quite difficult to obtain better performance. In order to address this problem, two novel deep models are proposed to extract more discriminative spatial-spectral features by exploiting the convolutional LSTM (ConvLSTM). By taking the data patch in a local sliding window as the input of each memory cell band by band, the 2-D extended architecture of LSTM is considered for building the spatial-spectral ConvLSTM 2-D neural network (SSCL2DNN) to model long-range dependencies in the spectral domain. To better preserve the intrinsic structure information of the hyperspectral data, the spatial-spectral ConvLSTM 3-D neural network (SSCL3DNN) is proposed by extending LSTM to the 3-D version for further improving the classification performance. The experiments, conducted on three commonly used HSI data sets, demonstrate that the proposed deep models have certain competitive advantages and can provide better classification performance than the other state-of-the-art approaches.
Mucus clearance is the primary defense mechanism that protects airways from inhaled infectious and toxic agents. In the current gel-on-liquid mucus clearance model, a mucus gel is propelled on top of ...a "watery" periciliary layer surrounding the cilia. However, this model fails to explain the formation of a distinct mucus layer in health or why mucus clearance fails in disease. We propose a gel-on-brush model in which the periciliary layer is occupied by membrane-spanning mucins and mucopolysaccharides densely tethered to the airway surface. This brush prevents mucus penetration into the periciliary space and causes mucus to form a distinct layer. The relative osmotic moduli of the mucus and periciliary brush layers explain both the stability of mucus clearance in health and its failure in airway disease.
Osteoarthritis (OA) is the most common joint disorder and affects approximately half of the aged population. Current treatments for OA are largely palliative until the articular cartilage has been ...deeply damaged and irreversible morphological changes appear. Thus, effective methods are needed for diagnosing and monitoring the progression of OA during its early stages when therapeutic drugs or biological agents are most likely to be effective. Various proteinases involved in articular cartilage degeneration in pre-OA conditions, which may represent the earliest reversible measurable changes, are considered diagnostic and therapeutic targets for early OA. Of these proteinases, matrix metalloproteinase 13 (MMP-13) has received the most attention, because it is a central node in the cartilage degradation network. In this review, we highlight the main MMP-13-related changes in OA chondrocytes, including alterations in the activity and expression level of MMP-13 by upstream regulatory factors, DNA methylation, various non-coding RNAs (ncRNAs), and autophagy. Because MMP-13 and its regulatory networks are suitable targets for the development of effective early treatment strategies for OA, we discuss the specific targets of MMP-13, including upstream regulatory proteins, DNA methylation, non-coding RNAs, and autophagy-related proteins of MMP-13, and their therapeutic potential to inhibit the development of OA. Moreover, the various entities mentioned in this review might be useful as early biomarkers and for personalized approaches to disease prevention and treatment by improving the phenotyping of early OA patients.
The detrimental effects of sleep insufficiency have been extensively explored. However, only a few studies have addressed this issue in adolescents. In the present study, we examined and compared the ...effects of 72 h paradoxical sleep deprivation (SD) on adolescent (5 weeks old) and adult (~12 weeks old) mice. Following 72 h of SD, induced by a modified multiple-platform method, mice were subjected to behavioral, histological and neurochemical examinations. In both adolescent and adult mice, SD adversely affected short-term memory in a novel object recognition test. Compared with normal-sleep controls, sleep-deprived adolescent mice had an increased density of excitatory synapses in the granule cells of the dentate gyrus, but no such pattern was observed in the adult group. The engulfment of postsynaptic components within the microglia after SD was reduced in adolescents but not in adults, suggesting an impaired microglia-mediated synaptic pruning in adolescent SD mice. Possible contributing factors included the decreases in CX3CR1, CD11b and P2Y12, closely associated with the synaptic pruning via microglial phagocytosis. In adult SD mice, microglia-associated inflammatory reactions were noted. In sum, sleep deprivation induces age-dependent microglial reactions in adolescent and adult mice, respectively; yet results in similar defects in short-term recognition memory. Sufficient sleep is indispensable for adolescents and adults.
•72 h sleep deprivation (SD) impairs short-term memory in adult & adolescent mice.•In adults, SD activates microglia and elevates proinflammatory cytokines.•In adolescents, SD defects synaptic pruning by reducing microglial phagocytosis.
We use scaling theory to derive the time dependence of the mean-square displacement ⟨Δr 2⟩ of a probe nanoparticle of size d experiencing thermal motion in polymer solutions and melts. Particles with ...size smaller than solution correlation length ξ undergo ordinary diffusion (⟨Δr 2(t)⟩ ∼ t) with diffusion coefficient similar to that in pure solvent. The motion of particles of intermediate size (ξ < d < a), where a is the tube diameter for entangled polymer liquids, is subdiffusive (⟨Δr 2(t)⟩ ∼ t 1/2) at short time scales since their motion is affected by subsections of polymer chains. At long time scales the motion of these particles is diffusive, and their diffusion coefficient is determined by the effective viscosity of a polymer liquid with chains of size comparable to the particle diameter d. The motion of particles larger than the tube diameter a at time scales shorter than the relaxation time τe of an entanglement strand is similar to the motion of particles of intermediate size. At longer time scales (t > τe) large particles (d > a) are trapped by entanglement mesh, and to move further they have to wait for the surrounding polymer chains to relax at the reptation time scale τrep. At longer times t > τrep, the motion of such large particles (d > a) is diffusive with diffusion coefficient determined by the bulk viscosity of the entangled polymer liquids. Our predictions are in agreement with the results of experiments and computer simulations.