An Empirical Study on the National Heterogeneity of High-End Manufacturing Technology Innovation and GVC’s Division of Labor Status

Although the relationship between technological innovation and the status of the global value chain’s (GVC) division of labor has been essentially affirmed by academia, the direct regression of all countries does not account for national differences pertaining to different economic development levels. This paper divides the countries selected for this study into developed and developing countries and then conducts empirical tests on two sample sets to explore the heterogeneity of technological innovation and GVC division of labor status. The results reveal the following: (1) in developed countries, the GVC division of labor status of high-end manufacturing is generally higher than that of developing countries; (2) in developed countries, the technological innovation of high-end manufacturing plays a significant role in promoting GVC’s division of labor, while developing countries have a significant inhibitory effect; and (3) staff input and financial developmental levels have significantly promoted GVC’s status in the division of labor. Earlier studies have shown that, in developing countries, technological innovation in high-end manufacturing industries does not fully serve the goal of exporting intermediate goods. This study’s conclusions offer a new method of explaining the nature of a given country, the logic of technological innovation, and the differences in the GVC division of labor status.

Parts-per-Object Count in Agricultural Images: Solving Phenotyping Problems via a Single Deep Neural Network

Solving many phenotyping problems involves not only automatic detection of objects in an image, but also counting the number of parts per object. We propose a solution in the form of a single deep network, tested for three agricultural datasets pertaining to bananas-per-bunch, spikelets-per-wheat-spike, and berries-per-grape-cluster. The suggested network incorporates object detection, object resizing, and part counting as modules in a single deep network, with several variants tested. The detection module is based on a Retina-Net architecture, whereas for the counting modules, two different architectures are examined: the first based on direct regression of the predicted count, and the other on explicit parts detection and counting. The results are promising, with the mean relative deviation between estimated and visible part count in the range of 9.2% to 11.5%. Further inference of count-based yield related statistics is considered. For banana bunches, the actual banana count (including occluded bananas) is inferred from the count of visible bananas. For spikelets-per-wheat-spike, robust estimation methods are employed to get the average spikelet count across the field, which is an effective yield estimator.

Leaf Counting: Fusing Network Components for Improved Accuracy

Leaf counting in potted plants is an important building block for estimating their health status and growth rate and has obtained increasing attention from the visual phenotyping community in recent years. Two novel deep learning approaches for visual leaf counting tasks are proposed, evaluated, and compared in this study. The first method performs counting via direct regression but using multiple image representation resolutions to attend leaves of multiple scales. The leaf count from multiple resolutions is fused using a novel technique to get the final count. The second method is detection with a regression model that counts the leaves after locating leaf center points and aggregating them. The algorithms are evaluated on the Leaf Counting Challenge (LCC) dataset of the Computer Vision Problems in Plant Phenotyping (CVPPP) conference 2017, and a new larger dataset of banana leaves. Experimental results show that both methods outperform previous CVPPP LCC challenge winners, based on the challenge evaluation metrics, and place this study as the state of the art in leaf counting. The detection with regression method is found to be preferable for larger datasets when the center-dot annotation is available, and it also enables leaf center localization with a 0.94 average precision. When such annotations are not available, the multiple scale regression model is a good option.

SDGH-Net: Ship Detection in Optical Remote Sensing Images Based on Gaussian Heatmap Regression

The ship detection task using optical remote sensing images is important for in maritime safety, port management and ship rescue. With the wide application of deep learning to remote sensing, a series of target detection algorithms, such as faster regions with convolution neural network feature (R-CNN) and You Only Look Once (YOLO), have been developed to detect ships in remote sensing images. These detection algorithms use fully connected layer direct regression to obtain coordinate points. Although training and forward speed are fast, they lack spatial generalization ability. To avoid the over-fitting problem that may arise from the fully connected layer, we propose a fully convolutional neural network, SDGH-Net, based on Gaussian heatmap regression. SDGH-Net uses an encoder–decoder structure to obtain the ship area feature map by direct regression. After simple post-processing, the ship polygon annotation can be obtained without non-maximum suppression (NMS) processing. To speed up model training, we added a batch normalization (BN) processing layer. To increase the receptive field while controlling the number of learning parameters, we introduced dilated convolution and added it at different rates to fuse the features of different scales. We tested the performance of our proposed method using a public ship dataset HRSC2016. The experimental results show that this method improves the recall rate of ships, and the F-measure is 85.05%, which surpasses all other methods we used for comparison.

DL-inferencing for 3D Cephalometric Landmarks Regression task using OpenVINO

In this paper, we evaluate the performance of the Intel Distribution of OpenVINO toolkit in practical solving of the problem of automatic three-dimensional Cephalometric analysis using deep learning methods. This year, the authors proposed an approach to the detection of cephalometric landmarks from CT-tomography data, which is resistant to skull deformities and use convolutional neural networks (CNN). Resistance to deformations is due to the initial detection of 4 points that are basic for the parameterization of the skull shape. The approach was explored on CNN for three architectures. A record regression accuracy in comparison with analogs was obtained. This paper evaluates the perfor- mance of decision making for the trained CNN-models at the inference stage. For a comparative study, the computing environments PyTorch and Intel Distribution of OpenVINO were selected, and 2 of 3 CNN architectures: based on VGG for regression of cephalometric landmarks and an Hourglass-based model, with the RexNext backbone for the land- marks heatmap regression. The experimental dataset was consist of 20 CT of patients with acquired craniomaxillofacial deformities and was in- clude pre- and post-operative CT scans whose format is 800x800x496 with voxel spacing of 0.2x0.2x0.2 mm. Using OpenVINO showed a great increase in performance over the PyTorch, with inference speedup from 13 to 16 times for a Direct Regression model and from 3.5 to 3.8 times for a more complex and precise Hourglass model.

PSXII-1 High-throughput Phenotyping of Rumen Microbial Contents Using Buccal Swabs

The rumen microbial community is the agent that allows cattle and other ruminants to process complex plant polymers into digestible fatty acids. Traditional methods to sample rumen microbes often involve labor-intensive stomach tubing, or invasive surgeries to access the rumen lumen via cannula ports, thereby limiting the number of animals that could be sampled in a specific study. In this study, we tested the viability of using buccal swabs as a proxy of the rumen microbial contents in a timecourse experiment on eight cannulated cows. Rumen contents and buccal swabs were collected at six equally spaced timepoints, with the first timepoint being 2 hours prior to feeding. Simpson diversity and Shannon evenness estimates of the microbial counts of each sample revealed that the first timepoint had the lowest diversity and highest evenness (Tukey HSD < 0.05) out of all other timepoints. Principal component analysis confirmed that the buccal swab samples from the first timepoint were the most similar to paired rumen samples taken at the same times. Using a Random Forest Classifier analysis, we estimated the Gini importance scores for individual microbial taxa as a proxy of their uniqueness to the rumen or oral environments of the cows. We identified 18 oral-only microbial taxa that are contaminants and could be removed from future comparisons using this method. Finally, we attempted to estimate the exact relative abundance of rumen microbial taxa from buccal swab samples using paired rumen-swab data in a Random Forest Regression model. The model was found to have moderate (~38%) accuracy in cross-validation studies. Our data suggests that buccal swabs can serve as fast and suitable proxies for rumen microbial contents of dairy cattle, but that additional factors must be measured to improve direct regression of results to those of the rumen.

Radiation Power Control of the Industrial CO2 Laser Excited by а Nonself-Sustained Glow Discharge by Changing the Frequency of Ionization Pulses

The article describes radiation power control of industrial CO2 lasers of Lantan series excited by а nonself-sustained glow discharge in the automatic mode. These lasers are closed-cycle fast gas-transport lasers excited by a nonself-sustained glow discharge with ionization by periodic-pulsed capacitively coupled auxiliary discharge. In this case, ionization and conductivity are provided by periodic-pulsed capacitively coupled discharge. The energy contribution to molecular oscillations is provided by the passage of the main discharge current through the plasma with electron density given by ionization. This permits easy laser power control, provides excellent optical homogeneity and stability of an active volume together with high laser efficiency. A system of a nonself-sustained glow discharge with ionization by periodic-pulsed capacitively coupled auxiliary discharge, the stages of creation and brief characteristics of the Lantan series lasers is presented. The method of controlling the power of laser radiation by changing the frequency of the ionization pulses is determined. This control method allows operating of the laser in continuous and in pulse-periodic modes with adjustable pulse ratio and pulse duration, and also provides switching from one mode to another. In the continuous mode, the radiation power is controlled by changing the frequency of ionization pulses, which are high voltage pulses with duration of 100 ns, given with the frequency of 1-5 kHz. Pulse-periodic radiation control is performed by modulating ionization pulses that consists of pulses being delivered in batches. The frequency of the pulses in a batch determines the radiation power in a pulse. The frequency of the batches following is the frequency of the pulse mode, and the length of the batch determines the pulses duration. Based on the experimental data, the dependence of the radiation power on the ionization pulses frequency was determined. An experimental system is presented and the measuring accuracy of the laser radiation power and the frequency of ionization pulses is determined. Data acquiring and processing of experimental results were performed using the NI 6008 USB data acquisition device in the LabVIEW programs of National Instruments. To study the dependence of the laser power on Мехатроника, автоматизация, управление, Том 21, № 4, 2020 231 the frequency of the ionization pulses, a regression analysis method was applied. Studies have shown that the dependence of the laser power on the ionization pulses frequency is linear in a wide range of parameters. The equation of the direct regression is calculated. The confidence estimates of the coefficients of the direct regression and the confidence estimates of the deviation of the theoretical direct regression from the empirical one are calculated with a confidence level of 95%.