Search

Pedestrian detection is a critical task for safety-critical systems, but detecting pedestrians is challenging in low-light and adverse weather conditions. Thermal images can be used to improve robustness by providing complementary information to RGB images. Previous studies have shown that multi-modal feature fusion using convolution operation can be effective, but such methods rely solely on local feature correlations, which can degrade the performance capabilities. To address this issue, we propose an attention-based novel fusion network, referred to as INSANet (INtra-INter Spectral Attention Network), that captures global intra- and inter-information. It consists of intra- and inter-spectral attention blocks that allow the model to learn mutual spectral relationships. Additionally, we identified an imbalance in the multispectral dataset caused by several factors and designed an augmentation strategy that mitigates concentrated distributions and enables the model to learn the diverse locations of pedestrians. Extensive experiments demonstrate the effectiveness of the proposed methods, which achieve state-of-the-art performance on the KAIST dataset and LLVIP dataset. Finally, we conduct a regional performance evaluation to demonstrate the effectiveness of our proposed network in various regions. [ABSTRACT FROM AUTHOR]

In this paper, multispectral pedestrian detection is mainly discussed, which can contribute to assigning human-aware properties to automated forklifts to prevent accidents, such as collisions, at an early stage. Since there was no multispectral pedestrian detection dataset in an intralogistics domain, we collected a dataset; the dataset employs a method that aligns image pairs with different domains, i.e. RGB and thermal, without the use of a cumbersome device such as a beam splitter, but rather by exploiting the disparity between RGB sensors and camera geometry. In addition, we propose a multispectral pedestrian detector called SSD 2.5D that can not only detect pedestrians but also estimate the distance between an automated forklift and workers. In extensive experiments, the performance of detection and centroid localization is validated with respect to evaluation metrics used in the driving car domain but with distinct categories, such as hazardous zone and warning zone, to make it more applicable to the intralogistics domain. [ABSTRACT FROM AUTHOR]

Plant disease detection is essential for optimizing agricultural productivity and crop quality. With the recent advent of deep learning and large-scale plant disease datasets, many studies have shown high performance of supervised learning-based plant disease detectors. However, these studies still have limitations due to two aspects. First, labeling cost and class imbalance problems remain challenging in supervised learning-based methods. Second, plant disease datasets are either unstructured or weakly-unstructured and the shapes of leaves and diseased areas on them are variable, rendering plant disease detection even more challenging. To overcome these limitations, we propose an instance-aware unsupervised plant disease detector, which leverages normalizing flows, a visual saliency map and positional encodings. A novel way to explicitly combine these methods is the proposed model, in which the focus is on reducing background noise. In addition, to better fit the model to the plant disease detection domain and to enhance feature representation, a feature extractor is pre-trained in a self-supervised learning manner using only unlabeled data. In our extensive experiments, it is shown that the proposed approach achieves state-of-the-art performance on widely-used datasets, such as BRACOL (Weakly-unstructured) and PlantVillage (Unstructured), regardless of whether the dataset is weakly-structured or unstructured. [ABSTRACT FROM AUTHOR]

Vertical farms are to be considered the future of agriculture given that they not only use space and resources efficiently but can also consistently produce large yields. Recently, artificial intelligence has been introduced for use in vertical farms to boost crop yields, and crop growth monitoring is an essential example of the type of automation necessary to manage a vertical farm system. Region of interest predictions are generally used to find crop regions from the color images captured by a camera for the monitoring of growth. However, most deep learning-based prediction approaches are associated with performance degradation issues in the event of high crop densities or when different types of crops are grown together. To address this problem, we introduce a novel method, termed pseudo crop mixing, a model training strategy that targets vertical farms. With a small amount of labeled crop data, the proposed method can achieve optimal performance. This is particularly advantageous for crops with a long growth period, and it also reduces the cost of constructing a dataset that must be frequently updated to support the various crops in existing systems. Additionally, the proposed method demonstrates robustness with new data that were not introduced during the learning process. This advantage can be used for vertical farms that can be efficiently installed and operated in a variety of environments, and because no transfer learning was required, the construction time for container-type vertical farms can be reduced. In experiments, we show that the proposed model achieved a performance of 76.9%, which is 12.5% better than the existing method with a dataset obtained from a container-type indoor vertical farm. Our codes and dataset will be available publicly. [ABSTRACT FROM AUTHOR]

Sap water is distributed and utilized through xylem conduits, which are vascular networks of inert pipes important for plant survival. Interestingly, plants can actively regulate water transport using ion-mediated responses and adapt to environmental changes. However, ionic effects on active water transport in vascular plants remain unclear. In this report, the interactive ionic effects on sap transport were systematically investigated for the first time by visualizing the uptake process of ionic solutions of different ion compositions (K+/Ca2+) using synchrotron X-ray and neutron imaging techniques. Ionic solutions with lower K+/Ca2+ ratios induced an increased sap flow rate in stems of Olea europaea L. and Laurus nobilis L. The different ascent rates of ionic solutions depending on K+/Ca2+ ratios at a fixed total concentration increases our understanding of ion-responsiveness in plants from a physicochemical standpoint. Based on these results, effective structural changes in the pit membrane were observed using varying ionic ratios of K+/Ca2+. The formation of electrostatically induced hydrodynamic layers and the ion-responsiveness of hydrogel structures based on Hofmeister series increase our understanding of the mechanism of ion-mediated sap flow control in plants. [ABSTRACT FROM AUTHOR]

The fabrication of gratings including metal deposition processes for highly neutron absorbing lines is a critical issue to achieve a good visibility of the grating-based phase imaging system. The source grating for a neutron Talbot-Lau interferometer is an array of Gadolinium (Gd) structures that are generally made by sputtering, photo-lithography, and chemical wet etching. However, it is very challenging to fabricate a Gd structure with sufficient neutron attenuation of approximately more than 20 μm using a conventional metal deposition method because of the slow Gd deposition rate, film stress, high material cost, and so on. In this article, we fabricated the source gratings for neutron Talbot-Lau interferometers by filling the silicon structure with Gadox particles. The new fabrication method allowed us a very stable and efficient way to achieve a much higher Gadox filled structure than a Gd film structure, and is even more suitable for thermal polychromatic neutrons, which are more difficult to stop than cold neutrons. The newly fabricated source gratings were tested at the polychromatic thermal neutron grating interferometer system of HANARO at the Korea Atomic Energy Research Institute, and the visibilities and images from the neutron phase imaging system with the new source gratings were compared with those fabricated by a Gd deposition method. [ABSTRACT FROM AUTHOR]

Abstract: Although titanosaurs represent one of the most diverse radiations of non-avian dinosaurs during the Cretaceous, our knowledge of their early developmental stages was restricted to the Auca Mahuevo (Argentina) embryos in ovo. Here, we present the first complete lithostrotia titanosaur embryo in ovo. The relatively small spherical 87.07 to 91.1mm egg was discovered at the Lower Cretaceous locality Algui Ulaan Tsav in Mongolia, and is, to date, the smallest positively identified titanosaur egg. Through taphonomic processes, the egg was transformed into a calcite geode at the bottom of which the embryonic bones settled down and are now partially exposed on the lower egg surface. Neutron tomography characterization reveals a fully developed embryo fossilized within a thin (7.6mm–8.6mm) calcite layer. EBSD, a SEM-based diffraction technique, which measures the complete crystallographic orientation of the crystal lattice from a submicron area on the sample surface, is used for the first time on an extinct dinosaur eggshell. Observations of the egg and its embryo combined with eggshell microcharacterizations suggest that this new embryo was a lithostrotia titanosaur with an intermediate robusticity index that shares a mosaic of skeletal characters with Diamantinasaurus matildae from Queensland (Australia) and the nemegtosaurid Rapetosaurus krausei (Madagascar) more than with any other titanosaurs. The Early Cretaceous age of Algui Ulaan Tsav implies that this specimen greatly predates the previously described lithostrotian titanosaurs from the Late Cretaceous sediments of Mongolia. In addition, the recognized amount of similar eggs that has been recovered during the last 70years at Algui Ulaan Tsav suggests that a well-established population of lithostrotian titanosaurs used this site as a nesting site. The combined observations provide an important addition to Mongolian fossil richness and alter our understanding of the paleodispersion of this sauropod group. It now appears that lithostrotian sauropods would have reached Mongolia during the Aptian–Albian, thus suggesting the existence of a passage before the complete separation of the Laurasian and Gondwanan continents. Possibly, the lithostrotian north and eastward migration could have occurred between North Africa, Spain, and the rest of Europe prior to its fragmentation in large islands during the Cretaceous, thus justifying the presence of lithostrotians in the Cretaceous Romanian Hateg Island. [Copyright &y& Elsevier]

Abstract: X-ray imaging detectors in combination with scintillator screens have been widely used in digital X-ray imaging applications. Gd2O2S:Tb was used as scintillation material for pixelated scintillator screens based on silicon substrates (wafer) with a micropore array of various dimensions fabricated using the photolithography and deep reactive ion etching (DRIE) process. The relative light output and the modulation transfer function (MTF) of each fabricated scintillator screen were measured by a cooled CCD and compared with those of Lanex screens. The spatial resolution of our scintillator screens was higher but their light outputs were lower than those of Lanex screen probably due to the loss of light at the wall surfaces. Therefore further treatment of the wall surface, such as reflective coating, seems necessary to compensate the light loss. [Copyright &y& Elsevier]

Various medicinal plants are threatened with extinction owing to their over-exploitation and the prevalence of soil borne pathogens. In this study, soils infected with root-rot pathogens, which prevent continuous-cropping, were treated with an electron beam. The level of soil-borne fungus was reduced to ≤0.01% by soil electron beam treatment without appreciable effects on the levels of antagonistic microorganism or on the physicochemical properties of the soil. The survival rate of 4-year-old plant was higher in electron beam-treated soil (81.0%) than in fumigated (62.5%), virgin (78%), or untreated-replanting soil (0%). Additionally, under various soils conditions, neutron tomography permitted the monitoring of plant health and the detection of root pathological changes over a period of 4-6 years by quantitatively measuring root water content in situ. These methods allow continual cropping on the same soil without pesticide treatment. This is a major step toward the environmentally friendly production of endangered therapeutic herbs.

Boiling heat transfer (BHT) is a particularly efficient heat transport method because of the latent heat associated with the process. However, the efficiency of BHT decreases significantly with increasing wall temperature when the critical heat flux (CHF) is reached. Graphene has received much recent research attention for applications in thermal engineering due to its large thermal conductivity. In this study, graphene films of various thicknesses were deposited on a heated surface, and enhancements of BHT and CHF were investigated via pool-boiling experiments. In contrast to the well-known surface effects, including improved wettability and liquid spreading due to micron- and nanometer-scale structures, nanometer-scale folded edges of graphene films provided a clue of BHT improvement and only the thermal conductivity of the graphene layer could explain the dependence of the CHF on the thickness. The large thermal conductivity of the graphene films inhibited the formation of hot spots, thereby increasing the CHF. Finally, the provided empirical model could be suitable for prediction of CHF.