Search

To address the challenges of missed detections caused by insufficient shape and texture features and blurred boundaries in existing detection methods, this paper introduces a novel moving vehicle detection approach for satellite videos. The proposed method leverages frame difference and convolution to effectively integrate spatiotemporal information. First, a frame difference module (FDM) is designed, combining frame difference and convolution. This module extracts motion features between adjacent frames using frame difference, refines them through backpropagation in the neural network, and integrates them with the current frame to compensate for the missing motion features in single-frame images. Next, the initial features are processed by a backbone network to further extract spatiotemporal feature information. The neck incorporates deformable convolution, which adaptively adjusts convolution kernel sampling positions, optimizing feature representation and enabling effective multiscale information integration. Additionally, shallow large-scale feature maps, which use smaller receptive fields to focus on small targets and reduce background interference, are fed into the detection head. To enhance small-target feature representation, a small-target self-reconstruction module (SR-TOD) is introduced between the neck and the detection head. Experiments using the Jilin-1 satellite video dataset demonstrate that the proposed method outperforms comparison models, significantly reducing missed detections caused by weak color and texture features and blurred boundaries. For the satellite-video moving vehicle detection task, this method achieves notable improvements, with an average F1-score increase of 3.9% and a per-frame processing speed enhancement of 7 s compared to the next best model, DSFNet. [ABSTRACT FROM AUTHOR]

The Royal Brisbane and Women's Hospital (RBWH) Swallowing and Nutrition Management Guidelines for Patients with Head and Neck Cancer were developed to enable evidence-based decision-making by the Head and Neck Multidisciplinary Team (H&N MDT) regarding enteral nutrition support options. The purpose of this study was to revalidate these guidelines in a cohort of patients receiving helical intensity-modulated radiotherapy (H-IMRT) compared to a historical cohort who received primarily 3D-conformal radiotherapy. Eligible patients attending the RBWH H&N MDT between 2013 and 2014 (n = 315) were assessed by the guidelines, with high-risk patients being recommended proactive gastrostomy tube placement. Data were collected on guideline adherence, gastrostomy tube insertions, the duration of enteral tube use and weight change. Sensitivity, specificity and positive predictive and negative predictive values were calculated and compared with the historical cohort. Overall guideline adherence was 84%, with 60% and 96% adherence to the high-risk and low-risk pathways, respectively. Seventy patients underwent proactive gastrostomy tube placement (n = 62 high-risk; n = 8 low-risk). Validation outcomes were sensitivity 73% (compared to 72%) and specificity 86% (compared to 96%). The guidelines yielded a high sensitivity and specificity, remaining valid in a cohort of patients treated with H-IMRT. Further studies are recommended to improve the sensitivity and understand the decrease in specificity in order to make ongoing guideline improvements. [ABSTRACT FROM AUTHOR]

The influence of solvent properties on ion generation by swab spray ionization was investigated. The ability of a variety of solvents of different relative permittivity, surface tension, and viscosity to form a stable and reproducible electrospray was examined. It is demonstrated that in swab spray ionization, a crucial balance between solvent composition, applied potential, and the solvent flow fed to the swab head must be maintained. The solvent composition was found to significantly affect the shape of the Taylor cone and the emerging cone jet, which eventually have an impact on the resulting ion yield. The results indicate that the relative permittivity of solvents measured under standard conditions is the main factor governing jet shaping, and consequently, the ionization efficacy. Short jets, which are required for maximum ion yield, were observed for solvents with relative permittivity εr higher than 25. Solvents exhibiting lower relative permittivity required the addition of 20% to 60% methanol to limit the jet length and to avoid the ineffective dripping pulsation. The observed effects were compared to conventional electrospray ionization and paper spray ionization. [ABSTRACT FROM AUTHOR]

The global trend of ore quality degradation, which is occurring in parallel with increased production volumes of primary metals, has led to the necessity of comprehensive processing of raw materials. This involves the concentration and recycling of metallurgical technogenic wastes, which were previously considered unprofitable. In particular, this means the introduction of dump tailings and sludges from holding ponds to concentration, as well as the feeding of matured concentrates, dump, and recycled slags, and in-house low-grade reverts to metallurgical processing. The feed and the product of its concentration demonstrate an increased content of the refractory spinel phase, which has a variable composition. A share of the technogenic feed in charge of head smelting units exceeds 20%. The majority of main smelting units employ autogenous processes, such as flash smelting, liquid bath smelting, and oxygen-flash smelting. The processing of technogenic low-heat feed in autogenous smelting units results in thermal imbalance, loss of melt homogeneity, and, consequently, disrupts the discharge of smelting products, potentially leading to complete shutdown of the unit. Such behavior is typical for both types of autogenous units operated in the Nornickel Polar Division, namely the Vanukov furnace and the flash smelting furnace. The issue of melt heterogeneity and the formation of refractory accretion at problem areas of the flash smelting furnace is widely known and sufficiently investigated in foreign literature as well as by the authors of this paper. However, further investigation is required for the Vanukov furnace, with a particular focus on the formation mechanism of the intermediate layer phase in the melt. This investigation should aim to develop effective mitigation strategies to address the adverse impact of this phase formation. [ABSTRACT FROM AUTHOR]

Existing point cloud object detection algorithms struggle to effectively capture spatial features across different scales, often resulting in inadequate responses to changes in object size and limited feature extraction capabilities, thereby affecting detection accuracy. To solve this problem, we present a point cloud object detection method based on multi-scale feature fusion of the original point cloud and projection, which aims to improve the multi-scale performance and completeness of feature extraction in point cloud object detection. First, we designed a 3D feature extraction module based on the 3D Swin Transformer. This module pre-processes the point cloud using a 3D Patch Partition approach and employs a self-attention mechanism within a 3D sliding window, along with a downsampling strategy, to effectively extract features at different scales. At the same time, we convert the 3D point cloud to a 2D image using projection technology and extract 2D features using the Swin Transformer. A 2D/3D feature fusion module is then built to integrate 2D and 3D features at the channel level through point-by-point addition and vector concatenation to improve feature completeness. Finally, the integrated feature maps are fed into the detection head to facilitate efficient object detection. Experimental results show that our method has improved the average precision of vehicle detection by 1.01% on the KITTI dataset over three levels of difficulty compared to Voxel-RCNN. In addition, visualization analyses show that our proposed algorithm also exhibits superior performance in object detection. [ABSTRACT FROM AUTHOR]

This article presents the findings of a study on the use of reactive nasogastric tube (NGT) feeding in patients with head and neck cancer (HNC) undergoing radiotherapy. The study found that most patients admitted for NGT feeding were receiving curative radiotherapy. The timing of NGT insertion varied, and the study suggests that streamlining access to NGT feeding could reduce hospital stays. The article also discusses the cost implications of NGT feeding and suggests that an outpatient NGT feeding service could be beneficial in initiating earlier enteral feeding and avoiding hospital admissions. [Extracted from the article]

Copyright of Chinese Journal of Applied Entomology is the property of Chinese Journal of Applied Entomology, Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Hog raisers of today consider feeds to be the most expensive component in hog enterprise. Thus, the study of the Comparative Effects of low-cost indigenous Feeds and Commercial Feeds on the Growth and Profitability of Native Pigs. Effects of indigenous feeds and commercial feeds on the growth and profitability of native pigs. This study was carried out to determine the effect of indigenous feeds and commercial feeds on the growth and profitability of native pigs. A total of 12 heads of piglets were distributed into 4 treatments with 3 replications. Results revealed that feeding native pigs with different mixtures of indigenous and commercial feeds has a significant effect on the initial weight; however, based on the results, it was T4 (100% commercial feeds) that showed a numerically heavier initial weight among all treatments fed with a different mixture of indigenous feeds, while in terms of weight gain, results showed that T1 was given 50% taro and 50% leftovers compared quantitatively to T4 fed 100% commercial feeds. Comparable differences were observed from T1-T3, where an average of 2,253.08 kg of feed per head was consumed. In other words, pigs consumed the maximum amount of feed given with a minimal feed leftover up to an average of 1.43 kg. Results revealed that feeds consumed by native pigs were numerically higher in T1-taro (50%) kitchen leftover (50%) followed by T3, T2 (80% Taro & 20% Kitchen leftover), while T4 with commercial feeds consumed 1,258.80 kg. In T3, 80% taro and 20% kitchen leftover, gave an average number of 8 piglets, followed by T2 with 7 piglets, T1 with 6 piglets, and T4 with 5 piglets, while in average weight, T4 has 0.8400 kg, T2 with 0.8267 kg, T3 with 0.6700 kg and T1 with 0.5667 kg, respectively. T3 have the highest profit of 11,707.60 pesos with Return on Investment (ROI) of 95.24%, followed by T2 with profit of 7,351.60 pesos and 53.86% ROI, T1 with 5,710.60 pesos and 46.47% ROI, and T4 with-43,180.40 pesos and ROI of -25.78%. [ABSTRACT FROM AUTHOR]

A post-processing technology is crucial in manufacturing complex curved surface parts as a bridge between CAD/CAM and machine manufacturing. Unlike the traditional five-axis machine tool, the five-axis additive–subtractive composite manufacturing machine tool combines additive and subtractive manufacturing. The previous studies focused on the tool vector control of five-axis laser additive manufacturing through the toolpath of five-axis helical milling. Compared with previous studies, this study combines additive and subtractive processes to examine the post-processing technology to coordinate and control the contour accuracy of forming parts in the composite manufacturing. First, the basic kinematic model of the five-axis additive–subtractive composite manufacturing machine tool is constructed on the basis of the inverse kinematic principle by assuming a laser nozzle coaxial powder feeding head. Second, the reasons for the incoordination of the two process coordinates are analyzed, and the criterion of process coordinate coordination is established on the basis of the five-axis general kinematic model. Finally, a special five-axis additive–subtractive composite post-processing software is developed based on a high-level language and verified by additive manufacturing, subtractive manufacturing, and process coordinate experiments. Experimental results show that the established composite post-processing model can effectively realize the contour accuracy of parts in the additive–subtractive composite manufacturing. This study provides new methods and ideas for the manufacturing accuracy and efficiency of five-axis additive-subtractive composite forming parts, which can broaden the application of thin-walled variable curvature parts through five-axis additive-subtractive composite post-processing technology. [ABSTRACT FROM AUTHOR]

The research of optimal compromise operation modes of five-column rectification unit, designed for isopropyl benzene extraction from benzene alkylation products by propylene, at variation of feed rate to the head tower is performed. We selected two optimization criteria in the form of productivity by outlet flow from the considered system of columns with the given concentration of isopropylbenzene and energy consumption for creation of steam flow in the columns. We varied feed flow rate to the head-column. Analysis of obtained results of the step-by-step approach for determining the optimal trade-off solution made it possible to establish that the multidimensional function of achieving a trade-off solution for the entire installation is unimodal. This result justifies the application of multidimensional extremum search methods to solve the problem of obtaining the optimal compromise solution. It has been established that the dependences of control variables on the supply flow rate to the head tower are linear when the optimum compromise modes are achieved. [ABSTRACT FROM AUTHOR]

The separation of fine coal has been widely discussed in the coal preparation industry due to its high economic potential. Dense medium cyclone (DMC) is the most efficient equipment available for fine coal separation. However, the industrial application of DMC is far from satisfactory due to operational difficulties and maintenance. In this research, particle settling behavior in a dense medium cyclone was analyzed for improved separation. The calculation result about feed pressure and separation lower limit, which fits the experimental data well, might be a guidance for industrial DMC design and operation. According to the calculation result, it is highly recommended that the separation lower limit be set at 0.2 mm rather than 0.1 mm, because the feed pressure head required for the latter (50 D) is three times higher than the former (15 D). [ABSTRACT FROM AUTHOR]

he purpose of this study was to explore the lowest loss rate of lysine and arginine content in the bran under different extrusion and expansion process parameters. With water content, feeding speed and head temperature as the test conditions, the results were analyzed by one-way ANOVA, and Design Expert 10.0.4 software was used to establish the results: Y¹=0.44+0.09A-3.74B+0.06C-0.02AB-0.09AC+ 0.03BC+0.07A² +0.05B² +0.10C² ; Y2=0.48-0.01A-2.50B+0.04C-0.01AB+2.49AC+0.07BC+0.12A² +0.12B² + 0.02C² (lysine content, Y1; arginine content, Y2; moisture content, A; feeding speed, B; mold head temperature, C). The results show that the process parameters were 30% water, 20 kg/h and head temperature of 110 ℃, and the lowest loss rates predicted by the secondary regression model were 43.80% and 48.40% respectively. After test verification, the lysine loss rate is 42.91% and the argi‐nine loss rate is 47.84%, and the measured value is close to the predicted value, in good agreement. It shows that the above two quadratic regression models have good predictive ability to better predict the actual rate of lysine and arginine loss in the process of bran expansion. [ABSTRACT FROM AUTHOR]

Many animal species can rapidly change their body colouration and patterning, but often the ecological drivers of such changes are unknown. Here, we explored dynamic colour change in the bluelined goatfish, Upeneichthys lineatus, a temperate marine teleost species. Upeneichthus lineatus can change in a matter of seconds, from a uniform white colour to display prominent, vertical, dark red stripes. Initial observations suggested that rapid colour change in U. lineatus was associated with feeding and may act as a signal to both conspecifics and heterospecifics that are frequently observed to follow feeding goatfish. Field observations of the colour and behaviour of individual U. lineatus were collected to (1) document the repertoire of behaviours that U. lineatus displays and categorise associated colour patterns; (2) quantify the speed of dynamic colour change; (3) establish the context in which U. lineatus changes colour and pattern; and (4) test whether the behaviour of follower fishes is influenced by colour patterning or specific behaviours of the focal goatfish. We found that U. lineatus changed colouration from white to the red banded pattern in less than 10 s. The key driver of rapid colour change in U. lineatus was feeding, particularly when the fish fed with its head buried in sediment. Conspecific followers were most likely to be white in colour and adopt searching behaviour, regardless of the focal fish colour or behaviour. Other species of follower fish spent significantly more time following U. lineatus that were displaying dark red stripes when searching or eating, implying the red stripes may be an interspecific signalling mechanism. Our findings indicate that rapid colour change in teleost fish may be used for social communication and may provide U. lineatus with increased protection from predation when feeding via a safety‐in‐numbers approach. [ABSTRACT FROM AUTHOR]

To solve the problems of complex structure and low efficiency in traditional brush head feeding system, the brush head feeding system was developed based on visual detection and positioning. The image of brush head was segmented by OTSU method, and the moment of image was calculated to obtain the gesture of brush head. The ellipse fitting of the brush head was carried out, and the center of the ellipse was taken as the position of brush head to judge the two sides of brush head by support vector machine. The robot trajectory was determined by 3-4-5 polynomial motion law to control the robot running smoothly, and the brush head feeding system was established. The test results show that the designed system can quickly and accurately realize the position determination of brush head, and the error is less than 0.3 mm. The system can work at the speed of 5 seconds per unit, which meets the demand of brush head feeding. [ABSTRACT FROM AUTHOR]

By damaging bark, mechanized harvesting deprives wood of its natural mechanical barrier. This study concerns the effect of this damage on the changes in moisture content that occur near the damaged areas of Scots pine (Pinus sylvestris L.). This study was carried out using 45 randomly selected 11 m long logged pine stems. Additionally, the effect of bark thickness on the depth of damage was measured. To determine the influence of the location of the damage and bark thickness on the wood moisture near the damaged region, wood samples were collected from two sections. The first was located one meter above the log's base, while the second was located one meter from the end of the log. Two increment cores were taken by a Pressler borer in each section: one from the damaged wood zone and the second from the undamaged wood zone. The average bark thickness one meter from the base of the log was 11.2 mm, which decreased to 1.8 mm in the samples taken one meter from the top of the log. The average depths of the damage caused by feed roller spikes in the two sections were 3.9 and 3.8 mm, respectively, indicating that there was no significant effect of bark thickness on the depth of the damage. The wood samples collected near the damaged wood zone (DW) had about 13 percent lower moisture content than those taken from the undamaged zone (UDW). This difference was statistically significant. We observed greater mean moisture content closer to the top of the log for both the samples taken near the damaged zone and the samples taken from the undamaged zone. One meter from the base of the log, the moisture content difference between the DW and UDW was 9.8 percentage points, which was statistically insignificant. By contrast, significantly lower moisture content (−16%) was observed in the DW one meter from the top of the log. [ABSTRACT FROM AUTHOR]

Fruit detection and recognition has an important impact on fruit and vegetable harvesting, yield prediction and growth information monitoring in the automation process of modern agriculture, and the actual complex environment of orchards poses some challenges for accurate fruit detection. In order to achieve accurate detection of green fruits in complex orchard environments, this paper proposes an accurate object detection method for green fruits based on optimized YOLOX_m. First, the model extracts features from the input image using the CSPDarkNet backbone network to obtain three effective feature layers at different scales. Then, these effective feature layers are fed into the feature fusion pyramid network for enhanced feature extraction, which combines feature information from different scales, and in this process, the Atrous spatial pyramid pooling (ASPP) module is used to increase the receptive field and enhance the network's ability to obtain multi-scale contextual information. Finally, the fused features are fed into the head prediction network for classification prediction and regression prediction. In addition, Varifocal loss is used tomitigate the negative impact of unbalanced distribution of positive and negative samples to obtain higher precision. The experimental results show that the model in this paper has improved on both apple and persimmon datasets, with the average precision (AP) reaching 64.3% and 74.7%, respectively. Compared with other models commonly used for detection, themodel approach in this study has a higher average precision and has improved in other performance metrics, which can provide a reference for the detection of other fruits and vegetables. [ABSTRACT FROM AUTHOR]

With the increasing number of cows per farmer in Japan, an automatic cow monitoring system is being introduced. One important aspect of such a system is the ability to identify individual cows and estimate their feeding time. In this study, we propose a method for achieving this goal through facial region analysis. We used a YOLO detector to extract the cow head region from video images captured during feeding with the head region cropped as a face region image. The face region image was used for cow identification and transfer learning was employed for identification. In the context of cow identification, transfer learning can be used to train a pre-existing deep neural network to recognize individual cows based on their unique physical characteristics, such as their head shape, markings, or ear tags. To estimate the time of feeding, we divided the feeding area into vertical strips for each cow and established a horizontal line just above the feeding materials to determine whether a cow was feeding or not by using Hough transform techniques. We tested our method using real-life data from a large farm, and the experimental results showed promise in achieving our objectives. This approach has the potential to diagnose diseases and movement disorders in cows and could provide valuable insights for farmers. [ABSTRACT FROM AUTHOR]

• Proxy token guided transformer-based baseline for vision-language tracking. • Dense annotated long-term vision-language tracking dataset. • Extensive experiments on a new long-term vision-language tracking dataset. Tracking by vision-language is an emergent topic. Previous researchers mainly adopt CNN and sequential models for video and language encoding, however, their methods are limited by poor generalization performance. To address this problem, this paper presents a novel vision-language tracking framework based on Transformer. Specifically, our proposed framework contains the image encoder, language encoder, cross-modal fusion module, and task-specific heads. We adopt the residual network and BERT for image and language embedding, respectively. More importantly, we propose a proxy token guided cross-modal fusion module based on the transformer network, which can link the vision and language features effectively and efficiently. The proxy token acts as a proxy for word embeddings and interacts with the visual feature. By absorbing vision information, the proxy token is used to modulate word embeddings and make them attend to the visual feature. Finally, we get the organically fused features via a dynamic modal aggregation method and feed them into the task-specific heads for tracking. Extensive experiments demonstrate that our method set new state-of-the-art on multiple language-assisted tracking datasets, including OTB-LANG, LaSOT, TNL2K, and a newly proposed Ref-LTB50 annotated with dense language specifications. Source code of this paper will be publicly available. [ABSTRACT FROM AUTHOR]

For small, arboreal marmoset monkeys, antipredator vigilance is essential to ensure survival. However, since antipredator vigilance is incompatible with some other behaviours, such as feeding in a head-down position, it has to be regulated efficiently within individuals and perhaps even at the group level. To investigate what drives individual differences in vigilance and feeding behaviour in common marmosets, Callithrix jacchus , and whether they adjust these behaviours to each other, we collected behavioural data on 14 marmosets in contexts where feeding and vigilance were mutually exclusive. The presence of infants resulted in higher overall vigilance and less feeding. Moreover, individuals increased their vigilance when the pair mate was feeding and thus could not be vigilant itself. They thus adjusted their own vigilance to the risk level of their pair mate, which resulted in a turn taking-like pattern of feeding and being vigilant in the pair. This flexibility of vigilance is consistent with marmosets' sensitivity to other's needs in another context, food sharing, when sharing is increased when food is more difficult to obtain. Together, these results suggest considerable awareness of the current state of group members in common marmosets. • Marmosets increased their vigilance when the partner was feeding head down. • This resulted in a turn-taking-like pattern of feeding and being vigilant. • Adults increased their vigilance and decreased feeding when infants were present. • Marmoset vigilance is flexible and takes others' risk level into account. [ABSTRACT FROM AUTHOR]

In robot-assisted oral surgery, the surgical tool needs to be fed into the target position to perform surgery. However, unmodeled extraoral and complex intraoral environments bring difficulties to motion planning. Meanwhile, the motion is operated manually by the surgeon, causing relatively limited accuracy as well as the risk of misoperation. Moreover, the random movements of the patient's head bring additional disturbance to the task. To achieve the task, a motion strategy based on a new conical virtual fixture (VF) was proposed. First, by preoperatively specifying a conical guiding cone as the VF, virtual repulsive forces were applied on the out-of-range end effector. Then, based on the two-point adjustment model and velocity conversion, the effect of VF was established to prevent the end-effector from exceeding the constraint region. Finally, a vision system corrects the guiding cone to compensate for the random movement of the patient's head to feed to a dynamic target. As an auxiliary framework for surgical operation, the proposed strategy has the advantages of safety, accuracy, and dynamic adaptability. Both simulations and experiments are conducted, verifying the feasibility of the proposed strategy. [ABSTRACT FROM AUTHOR]

Purpose: Automatic segmentation of colon, small intestine, and duodenum is a challenging task because of the great variability in the scale of the target organs. Multi‐scale features are the key to alleviating this problem. Previous works focused on extracting discriminative multi‐scale features through a hierarchical structure. Instead, the purpose of this work is to exploit these powerful multi‐scale features more efficiently. Methods: A Scale Attention Module (SAM) was proposed to recalibrate multi‐scale features by explicitly modeling their importance score adaptively. The SAM was introduced into the segmentation model to construct the Scale Attention Network (SANet). The multi‐scale features extracted from the encoder were first re‐extracted to obtain more specific multi‐scale features. Then the SAM was applied to recalibrate the features. Specifically, for the feature of each scale, a summation of Global Average Pooling and Global Max Pooling was used to create scale‐wise feature representations. According to the representations, a lightweight network was used to generate the importance score of each scale. The features were recalibrated based on the scores, and a simple pixel‐by‐pixel summation was used to fuse the multi‐scale features. The fused multi‐scale feature was fed into a segmentation head to complete the task. Results: The models were evaluated using fivefold cross‐validation on 70 upper abdominal computed tomography scans of patients in a volume manner. The results showed that SANet could effectively alleviate the scale‐variability problem and achieve better performance compared with UNet, Attention UNet, UNet++, Deeplabv3p, and CascadedUNet. The Dice similarity coefficients (DSCs) of colon, small intestine, and duodenum were (84.06 ± 3.66)%, (76.79 ± 5.12)%, and (61.68 ± 4.32)%, respectively. The HD95 were (7.51 ± 2.45) mm, (11.08 ± 2.45) mm, and (12.21 ± 1.95) mm, respectively. The values of relative volume difference were (3.4 ± 0.8)%, (11.6 ± 11.81)%, and (6.2 ± 3.71)%, respectively. The values of center‐of‐mass distance were 7.85 ± 2.82, 9.89 ± 2.70, and 9.94 ± 1.58, respectively. Compared with other attention modules and multi‐scale feature exploitation approaches, SAM could obtain a 0.83–2.71 points improvement in terms of DSC with a comparable or even less number of parameters. The extensive experiments confirmed the effectiveness of SAM. Conclusions: The SANet can efficiently exploit multi‐scale features to alleviate the scale‐variability problem and improve the segmentation performance on colon, small intestine, and duodenum of the upper abdomen. [ABSTRACT FROM AUTHOR]

Detecting accurate 3D bounding boxes of the object from point clouds is a major task in autonomous driving perception. At present, the anchor-based or anchor-free models that use LiDAR point clouds for 3D object detection use the center assigner strategy to infer the 3D bounding boxes. However, in the real-world scene, due to the occlusions and the effective detection range of the LiDAR system, only part of the object surface can be covered by the collected point clouds, and there are no measured 3D points corresponding to the physical object center. Obtaining the object by aggregating the incomplete surface point clouds will bring a loss of accuracy in direction and dimension estimation. To address this problem, we propose a corner-guided anchor-free single-stage 3D object detection model (CG-SSD). Firstly, the point clouds within a single frame are assigned to regular 3D grids. 3D sparse convolution backbone network composed of residual layers and sub-manifold sparse convolutional layers are used to construct bird's eye view (BEV) features for further deeper feature mining by a lite U-shaped network; Secondly, a novel corner-guided auxiliary module (CGAM) with adaptive corner classification algorithm is proposed to incorporate corner supervision signals into the neural network. CGAM is explicitly designed and trained to estimate locations of partially visible and invisible corners to obtain a more accurate object feature representation, especially for small or partial occluded objects; Finally, the deep features from both the backbone networks and CGAM module are concatenated and fed into the head module to predict the classification and 3D bounding boxes of the objects in the scene. The experiments demonstrate that CG-SSD achieves the state-of-art performance on the ONCE benchmark for supervised 3D object detection using single frame point cloud data, with 62.77% mAP. Additionally, the experiments on ONCE and Waymo Open Dataset show that CGAM can be extended to most anchor-based models which use the BEV feature to detect objects, as a plug-in and bring +1.17%∼+14.23% AP improvement. The code is available at https://github.com/mrqrs/CG-SSD. [ABSTRACT FROM AUTHOR]

Fruit object detection is crucial for automatic harvesting systems, serving applications such as orchard yield measurement and fruit harvesting. In order to achieve fast recognition and localization of green apples and meet the real-time working requirements of the vision system of harvesting robots, a fast optimized Foveabox detection model (Fast-FDM) is proposed. Fast-FDM uses an optimized form of anchor-free Foveabox to accurately and efficiently detect green apples in harvesting environments. Specifically, the EfficientNetV2-S with fast training and small size is used as the backbone network, a weighted bi-directional feature pyramid network (BiFPN) is employed as the feature extraction network to fuse multi-scale features easily and fast, and then the fused features are fed to the fovea head prediction network for the classification and bounding box prediction. Furthermore, an adaptive training sample selection (ATSS) method is adopted to directly select positive and negative samples, allowing green fruits of different scales to obtain higher recall and achieve more accurate green apple detection. Experimental results show that the proposed Fast-FDM realizes a mean average precision (mAP) of 62.3% for green apple detection using fewer parameters and floating point of operations (FLOPs), achieving better trade-offs between accuracy and detection efficiency. [ABSTRACT FROM AUTHOR]

Purpose: In this paper, continuous glass tow preg-reinforced acrylonitrile butadiene styrene (ABS) composites were fabricated by using a 3D printing method, and the purpose of this study is the investigation of the fiber preimpregnation effect on the mechanical behavior of these composites. In addition, a simple theoretical approach (mixture law), which considers the elastic behavior of reinforced composites and a numerical simulation method based on finite element method (FEM), was used to predict the tensile stress–strain behavior of ABS/glass tow preg composites in the elastic region. Design/methodology/approach: Different groups of preimpregnated glass tows with various ABS amounts (named 2%, 10%, 20% and 30%) were prepared by the solution impregnation method. Then, preimpregnated glass tows (prepregs or tow-pregs) were fed into the printer head along with the polymeric ABS filament to print the composites. The tensile, flexural and short beam tests were conducted to evaluate the mechanical properties of the printed composites. Findings: The first result of using tow-pregs instead of dry tows in continuous fiber 3D printing is much easier printing, printability improvement and the possibility of printing layers with low thickness, which can further increase the mechanical properties. The mechanical test results showed all of the glass prepregs improve strength and modulus in the tensile, three-point bending and short beam tests compared with neat ABS specimens, but statistical analysis showed that ABS weight percentage in the prepregs had no significant effect on the mechanical strength of composites except for the tensile modulus. Samples containing 2%-prepreg (minimum ABS amount in the tow-pregs) showed a significant improvement in tensile modulus. In the simulation section, good agreement is obtained between the model predictions and experimental tensile results. The results show that an acceptable deviation (14%) exists between the experimental and predicted value of elastic modulus by the numerical model. Originality/value: To the best of the authors' knowledge, this is the first study showing the effects of ABS weight percentage in prepregs on the mechanical properties of 3D printed continuous fiber-reinforced composites and predicting the mechanical behavior of 3D printed composites by numerical simulation method. [ABSTRACT FROM AUTHOR]

Form–function studies have established a strong link between dental morphology and the mechanical properties of food items, with animals evolving tooth shapes theoretically ideal for their diets. However, information on how teeth perform under natural conditions is rare, which limits the understanding of how dental morphology influences dietary ecology and niche partitioning within animal communities.Free‐tailed bats (Chiroptera: Molossidae) are a diverse clade of aerial insectivorous mammals that exhibit an outstanding variation in size and craniodental traits, which have been directly related to ecological segregation among sympatric species.We investigate the mechanisms that allow functional dietary specialization and trophic segregation among sympatric free‐tailed bat species inhabiting a Neotropical forest. To do so, we coupled data on 3D dental topographic metrics, head and skull dimensions, field‐collected feeding performance and dietary hardness measurements.We found that evolved differences in molar topography and skull size vary in tandem with the mechanical demands of prey items naturally consumed by sympatric molossid species. This may be explained by feeding performance capabilities resulting from both molar shape and the overall size of the feeding apparatus, which seem to allow efficient processing of prey items with specific mechanical properties. For instance, smaller bats with higher molar topographic values (sharper, more complex molars) and more gracile heads mainly feed on softer insects, whereas bigger bats with lower molar topographic values (blunter, less complex molars) and more robust heads mostly feed on tougher insects. Species with a broader range of sizes, craniodental morphologies and insect hardness are also present in the community.These results illustrate how the morphology and size of feeding structures, and how they perform, may facilitate trophic segregation among sympatric insectivorous bats. Similar mechanisms may help structure other communities of insectivorous mammals, therefore the approaches presented here could be used to generate a better understanding of the ecomorphological traits and processes that underlie their diversity. Read the free Plain Language Summary for this article on the Journal blog. [ABSTRACT FROM AUTHOR]

An experiment was conducted to investigate the effect of screw speed, feeding head speed, and the nature and type of granular material on the productivity and power consumption of a screw conveyor, and then determine the optimal speed. This experiment contains two parts. Part one was designed with 11 levels of screw speed (100-600r/min, interval 50r/min), three levels of feeding head speed (60,80 and 100r/min), and two kinds of granular materials (fine sand and soybeans). Part two was designed with 11 levels of screw speed (20-140r/min, interval 20r/min), three levels of feeding head speed (300,400 and 500r/min), and two kinds of granular materials (fine sand and soybeans). The results indicated that with the increase of the screw speed, the conveying performance will not always get better; with the rise of the feeding head speed, the conveying performance will get better. Due to the higher bulk density of sand, the productivity of conveying sand is higher for granular materials. Finally, for sand, the result indicates when the screw speed is 400r/min, and the feeding head speed is 120r/min, the conveying performance of sand determined by the productivity and power consumption is the best; for soybean, when the screw speed is 300r/min, and the feeding head speed is 140r/min, the conveying performance is the best. [ABSTRACT FROM AUTHOR]

Background: Nutritional care in patients with a severe head injury is of utmost importance, as nutritional deficiency can affect the outcomes in these patients. The purpose of this study was to assess the feasibility, and an outcome of performing bedside percutaneous endoscopic gastrostomy in severe head injury patients. This study was conducted in a tertiary care institute. Patients with head injury with a Glasgow coma scale score ≤ 8 were included. Among 50 patients, 25 patients underwent percutaneous endoscopic gastrostomy and 25 patients underwent open feeding gastrostomy, within 72 h to up to 3 weeks from admission. Results: The commonest subset in the study population was between18 and 30 years age group. Among 50 patients 43 (86%) were men and 7 (14%) were women with the men to women ratio being 6:1. The mean operative time for performing PEG was 12.04 SD 3.72 min, whereas for open feeding gastrostomy was 96.8 SD 9.98 min which was statistically significant (P = 0.0001). The complication rate for the PEG subgroup was 24% and that for feeding gastrostomy was 56% which was statistically significant (P = 0.02). Conclusions: Percutaneous endoscopic gastrostomy is a minimally invasive and time-saving procedure, which is superior to open feeding gastrostomy in terms of a prerequisite of anaesthesia, operating room, and few post-operative complications. [ABSTRACT FROM AUTHOR]

‘Miwoo’ is a mid-late maturing, high-biomass-yielding rice (Oryza sativa L.) cultivar with high feed value and resistance to multiple diseases and insect pests. It was developed for use as whole crop silage (WCS) from a cross between ‘Suweon519’, a line with high biomass yield and multiple disease resistance, and ‘Suweon518’, a medium flowering, high yielding Tong-il type line, which is resistant to blast (BL), bacterial blight (BB) (race K1), rice stripe virus (RSV) and brown planthopper (BPH). In the central plain region of Suwon, ‘Miwoo’ cultivar exhibited a growth period of approximately 122 days from seeding to heading, culm length of 99 cm, panicle length of 28 cm, 12 panicles per hill, 143 spikelets per panicle, and 1,000-grain weight of 25.9 g as brown rice. This mid-late-flowering rice variety was resistant to lodging during the yellowing ripening stage, an optimal period for harvesting. In addition, ‘Miwoo’ was strongly resistant to leaf and neck blast, BB (race K1, K2, K3, K3a), RSV, BPH and small BPH. Its average dry matter yield over 3 years reached 19.9 MT/ha, which is 36% higher than that of ‘Nokyang’. ‘Miwoo’ is a high-amylose cultivar with 70.7% total digestible nutrients (similar to that of Nokyang) and grows well on the central and southern plains of the Korean Peninsula. It should be harvested 30 days after heading to improve its feed value and digestion rate in livestock. [ABSTRACT FROM AUTHOR]

Extracting the leaves has been one of the most fundamental researches for the phenotypic traits of plants. However, it is difficult to further improve the accuracy of leaves segmentation, due to the mutual occlusion between leaves in rosette plants. It is still lacking in the leaf edge characteristics, particularly for the small target of young leaves. In this study, an improved instance segmentation benchmark Mask R-CNN model was proposed to realize an accurate segmentation for the plant leaves. The Cascade R-CNN was also introduced to generate the selection of the proposal boxes for the region proposal network. The boxes were first sent into the Head1 convolutional network with a threshold of 0.5, and then the detection was input into the Head 2 convolutional network with a threshold of 0.6, and finally, the detection was fed into the Head 3 convolutional network with a threshold of 0.7. Three cascaded networks were used to gradually increase the threshold, where the output of the previous network was applied to the higher threshold of the next network, indicating the high quality of the detection branch. After the segmentation branch full convolutional layer (FCN), an attention mechanism (SeNet) and a two-layer 3×3 convolution module were added to weigh the features of the segmentation branch, and further to extract the leaf edge segmentation. The segmentation of multi-scale leaves was also adopted for the leaves of different sizes in an image in the test process. The original and enlarged images were input into the model at the same time for the leaves segmentation. The optimal target on multiple scales was achieved using Non-Maximum Suppression (NMS). The ResNeXt101 was selected as the backbone network to extract the characteristics of leaves. An improved model of deep learning was also utilized at the same time. The image enhancement included random mirroring, rotation, enlargement, and size adjustment. There was no longer requiring for the special function library support, indicating the simplified data enhancement. The improved model had effectively reduced the blade occlusion, edge blur, and small target in the blade segmentation. Correspondingly, the Symmetric Best Dice (SBD) was 90.3% in the CVPPP Leaf Segmentation Challenge (LSC), which was 2.3 percentage points higher than that in Mask R-CNN, particularly than 1 percentage points than before. The improved neural network model was suitable for the leaf segmentation. The finding can provide a more accurate leaf segmentation for the plant phenotype using image processing, indicating a high application value in the research field of plant leaves. [ABSTRACT FROM AUTHOR]

Purpose: We assessed current policies and perceptions of enteral tube feeding in head and neck cancer (HNC) patients undergoing chemoradiotherapy (CRT). A web-based survey was used to compare the timing and method of tube feeding within and between the five Nordic countries, covering a population of 27 million. Methods: From each of the 21 Nordic university hospitals, one oncologist and one otorhinolaryngology—head and neck (ORL-HN) surgeon responded to a survey. The respondents were asked whether tube insertion before the onset of CRT, or during CRT as required (prophylactic vs. reactive) was preferred. The use of a pretreatment nutritional screening tool and the choice of feeding route (nasogastric vs. gastrostomy tube) were assessed. In total, we analyzed responses from 21 oncologists and 21 ORL-HN surgeons. Results: A tendency was observed towards decreasing the use of a PEG tube. Of the 21 university hospitals, only 2 (10%) reported using a prophylactic PEG tube in over half of HNC patients undergoing definitive CRT. The preferred method for reactive tube feeding was by a NG tube in 14 of 21 (67%), and by a PEG in 7 of 21 (33%). In general, both oncologists and ORL-HN surgeons were content with their current policy. Conclusions: The practices for enteral tube feeding in HNC management vary within and between the Nordic countries. We suggest that unified protocols for tube feeding should be developed for this patient population. [ABSTRACT FROM AUTHOR]

Members of the subfamily Heterocongrinae (Congridae) are a peculiar group of anguilliform eels that construct sandy borrows, form large colonies, and are popularly recognized as garden eels. They live with most of their bodies inside self-constructed borrows exposing their heads and trunk to feed on zooplankton, preferably copepods, that are brought passively by currents. As plankton feeders there was a suspicion that their branchial skeleton would have structures that could aid in the filtering process, such as highly developed or modified branchial rakers, which are observed in other suspension-feeding fishes, such as anchovies and sardines. Branchial rakers, however, were considered to be absent across Anguilliformes (except for Protanguilla). Nonetheless, specimens that were examined using clearing and staining and computed tomography showed, in all cases, branchial rakers associated with their gill arches. Heterocongrines have branchial rakers across their first to fourth branchial arches. These rakers are conical and apparently unossified, but further studies are necessary to attest its degree of ossification or its complete absence. Their pharyngeal tooth plates are reduced, a condition that may reflect their preference for smaller food items. Additionally, they may use crossflow filtering to feed, although detailed studies are necessary to clarify if hydrosol sieving may also aid in food capture. Furthermore, the present study proposes that the presence of branchial rakers should be better investigated in Anguilliformes with similar feeding habits as heterocongrines, considering that these structures may be more widespread within the group than previously considered. • Members of Heterocongrinae bear unossified branchial rakers in their branchial arches. • The presence of branchial rakers reflects the plankton feeding diet of heterocongrines. • Branchial rakers may have a more widespread distribution across Anguilliformes. [ABSTRACT FROM AUTHOR]

This article considers the trunk lines, connected with the feeder region as the external environment for the system - feeder lines (its subsystems are container flows, ports and container ships). Hub ports are considered to be the boundary of the system and the external environment. Thus, mathematical models were simulated for the two levels. At the first level, container flows are determined (based on statistical data), which enter and exit the hub. At the second level, the mathematical model allows to find for the feeder line: volumes of container flows, the total capacity of the ships which serves the cargo, and parts of the distribution of this capacity between feeder ports. [ABSTRACT FROM AUTHOR]

We propose a theory of head displacement that replaces traditional Head Movement and Lowering with a single syntactic operation of Generalized Head Movement. We argue that upward and downward head displacement have the same syntactic properties: cyclicity, Mirror Principle effects, feeding upward head displacement, and being blocked in the same syntactic configurations. We also study the interaction of head displacement and other syntactic operations, arguing that claimed differences between upward and downward displacement are either spurious or follow directly from our account. Finally, we show that our theory correctly predicts the attested crosslinguistic variation in verb and inflection doubling in predicate clefts. [ABSTRACT FROM AUTHOR]