Shot Training Research Articles

A general-purpose organic gel computer that learns by itself

To build energy minimized superstructures, self-assembling molecules explore astronomical options, colliding ∼109 molecules s−1. Thus far, no computer has used it fully to optimize choices and execute advanced computational theories only by synthesizing supramolecules. To realize it, first, we remotely re-wrote the problem in a language that supramolecular synthesis comprehends. Then, all-chemical neural network synthesizes one helical nanowire for one periodic event. These nanowires self-assemble into gel fibers mapping intricate relations between periodic events in any-data-type, the output is read instantly from optical hologram. Problem-wise, self-assembling layers or neural network depth is optimized to chemically simulate theories discovering invariants for learning. Subsequently, synthesis alone solves classification, feature learning problems instantly with single shot training. Reusable gel begins general-purpose computing that would chemically invent suitable models for problem-specific unsupervised learning. Irrespective of complexity, keeping fixed computing time and power, gel promises a toxic-hardware-free world. One sentence summary: fractally coupled deep learning networks revisits Rosenblatt’s 1950s theorem on deep learning network.

Multi-view and Multi-level network for fault diagnosis accommodating feature transferability

Various deep transfer learning solutions have been developed for machine fault diagnosis. The existing solutions mainly focus on domain adaptation by minimizing the data distribution discrepancy with certain metric, which emphasize the common features embedded in the data cross domains and neglect the unique features toward health condition classification in one specific domain. In these solutions, all the data for training have been forced to align in a common feature space and all the features for domain adaptation have been treated equally. However, there might exist domain specific features which are not appropriate for transfer but may contain essential information for classification in specific domain. In addition, due to the difficulty of collecting machine fault data, the number of machine fault samples is usually quite small or even zero. The traditional deep network structures and the training strategy are not the optimal choice in this occasion. To address these problems, a novel multi-view and multi-level network (MMNet) for fault diagnosis is developed. In MMNet, two network channels have been respectively constructed for cross domain common feature and domain specific feature learning to provide multi-view features. This architecture could implicitly differentiate the common features cross domains and the specific features only in one domain. In the channel of domain specific feature, a domain classifier and fault classifier are combined to learn the domain specific features. Multiple kernel maximum mean discrepancy (MK-MMD) is imposed on multiple layers of the common feature channel to implement domain adaptation and extract cross domain common features. The domain classification and fault classification together form a multi-level classification scheme. A classic few shot learning architecture with two modules respectively for feature extraction and relation computation is adopted as the backbone network. The relation score based classification mechanism enables zero shot fault classification in the target domain. Episode based few shot training strategy is employed to enhance the performance of MMNet with few labeled training data. Extensive experiments have demonstrated the state-of-the-art performance of MMNet on the involved transfer tasks.

Basketball jump shot technique design for high school athletes: Training method development

This study aims to improve the results of basketball jump shot training through a study of the development of training methods. The research method used is development research. Development research that refers to the development model of Borg and Gall, namely: 1) preliminary study, 2) early development model, 3) expert evaluation, 4) small-scale trial, 5) product revision, 6) group trial, 7) refine the model, 8) define the final model. Data are collected using a questionnaire obtained from the evaluation of basketball experts and coaches, then tested with a small group of 12 athletes and a large-scale trial of 24 athletes. The data are the results of an assessment of product quality, suggestions for product improvement, and the results of filling out questionnaires by athletes. The data analysis technique used is descriptive percentage to reveal the athlete's assessment after using the product. The results of the validation of the development of training methods from basketball experts obtain an average percentage of 94.5%. While the validation of the development of training methods from basketball coaches obtain an average percentage of 96.2%. The results of the questionnaire that is filled out by 12 athletes are 21.7% who answering difficult, 46.7% answering moderately, and 31.6% answering easy. The results of the analysis of small-scale trial data from 12 athletes obtain an average percentage of 20.83% answering difficult, 50% answering moderate, and 29.16% answering easy. The results of the analysis of large-scale trial data from 24 athletes obtain an average percentage of 6.7% who answering difficult, 19.1% answering moderately, and 74.1% answering easy. From the available data, it can be concluded that the development of the basketball jump shot training method can be used for basketball athletes from SMA Negeri 1 Majalengka.

Modified generalized neo-fuzzy system with combined online fast learning in medical diagnostic task for situations of information deficit

<abstract> <p>In the paper, we propose the modified generalized neo-fuzzy system. It is designed to solve the pattern-image recognition task by working with data that are fed to the system in the image form. The neo-fuzzy system can work with small training datasets, where classes can overlap in a features space. The core of the system under consideration is a modification of multidimensional generalized neuro-fuzzy neuron with an additional softmax activation function in the output layer instead of the defuzzification layer and quartic-kernel functions as membership ones. The learning procedure of the system combined cross-entropy criterion optimization using a matrix version of the optimal by speed Kaczmarz-Widrow-Hoff algorithm with the additional filtering (smoothing) properties. In comparison to the well-known systems, the modified neo-fuzzy one provides both numerical and computational implementation simplicity. The computational experiments have proved the effectiveness of the modified generalized neo-fuzzy-neuron, including the situation with shot training datasets.</p> </abstract>

ANALYSIS OF DIFFERENT INJURIES OF BASKETBALL PLAYERS BASED ON SURFACE ELECTROMYOGRAPHY

ABSTRACT Assessing the performance of basketball players is very important in the implementation of technical training programs. The application of electromyography (EMG) in basketball players is still relatively small. The evaluation of athletes’ muscle state index by EMG can guide sports training. This study used surface electromyography to test and compare EMG data, analyze muscle discharge timing, contribution rate and integral EMG value of the turning movement, aiming to explore the prevention mechanism of different types of injury affecting basketball players. The synchronous measurement model of EMG signal and plantar pressure curve in jump shot process of basketball players is established to deepen coaches’ understanding of jump shot movement rules and guide athletes to conduct scientific jump shot training. The research shows that the incidence of knee joint injury is relatively high in basketball players. Patella strain and softening, knee meniscus injury and cruciate ligament injury are the main types of injury, and there is no difference between genders.

Seg2pix: Few Shot Training Line Art Colorization with Segmented Image Data

There are various challenging issues in automating line art colorization. In this paper, we propose a GAN approach incorporating semantic segmentation image data. Our GAN-based method, named Seg2pix, can automatically generate high quality colorized images, aiming at computerizing one of the most tedious and repetitive jobs performed by coloring workers in the webtoon industry. The network structure of Seg2pix is mostly a modification of the architecture of Pix2pix, which is a convolution-based generative adversarial network for image-to-image translation. Through this method, we can generate high quality colorized images of a particular character with only a few training data. Seg2pix is designed to reproduce a segmented image, which becomes the suggestion data for line art colorization. The segmented image is automatically generated through a generative network with a line art image and a segmentation ground truth. In the next step, this generative network creates a colorized image from the line art and segmented image, which is generated from the former step of the generative network. To summarize, only one line art image is required for testing the generative model, and an original colorized image and segmented image are additionally required as the ground truth for training the model. These generations of the segmented image and colorized image proceed by an end-to-end method sharing the same loss functions. By using this method, we produce better qualitative results for automatic colorization of a particular character’s line art. This improvement can also be measured by quantitative results with Learned Perceptual Image Patch Similarity (LPIPS) comparison. We believe this may help artists exercise their creative expertise mainly in the area where computerization is not yet capable.

Supervised learning with a quantum classifier using multi-level systems

We propose a quantum classifier, which can classify data under the supervised learning scheme using a quantum feature space. The input feature vectors are encoded in a single qu$N$it (a $N$ level quantum system), as opposed to more commonly used entangled multi-qubit systems. For training we use the much used quantum variational algorithm -- a hybrid quantum-classical algorithm -- in which the forward part of the computation is performed on a quantum hardware whereas the feedback part is carried out on a classical computer. We introduce "single shot training" in our scheme, with all input samples belonging to the same class being used to train the classifier simultaneously. This significantly speeds up the training procedure and provides an advantage over classical machine learning classifiers. We demonstrate successful classification of popular benchmark datasets with our quantum classifier and compare its performance with respect to some classical machine learning classifiers. We also show that the number of training parameters in our classifier is significantly less than the classical classifiers.

The effect of shooting training on the development of the shot hit rate for basketball players

The aim of this study is to investigate the effect of shooting training for young basketball players and the development of shot percentage of special shot training. 30 male students-athletes participated in this study voluntarily participated in College Teams (age= 14,82 ± 1,0; height= 183,44 ± 6,13; body weight= 68,06 ± 5,38). Subjects were divided into 2 groups, consisting of continuous shooting training (experiment) (n = 15) and general basketball training group (control) (n = 15).Both training groups were subjected to a training program for 10 weeks, 4 days in a week. In this study, 60 minutes of normal basketball training plus shot training and 100-110 minutes of normal basketball training were applied to the control group. During the 10-week period, two-points jump shot (2000 pieces), three-points jump shot (2000 pieces) and in the Zig-Zag run (two-points jump shot-1000 pieces , three-points jump shot-1000 pieces) and totally 6000 shots used as moving. There was no statistically significant difference in the pre-test results between the control and the experimental group in the study, and 2-points, 3-points and zig-zag between the control and the experimental group. However there is a statistically significant difference in all parameters in the final test. In the study, 2-points, 3-points and zig-zag drills 2 and 3-points shots pre-test and post-test results were found to be statistically significant in the experimental group. As a result; general basketball training has shown little improvement in young basketball players' shooting performance, but it has been proven that the long shot training with accurate shot technique training has significantly improved shot performance.

Explore Efficient Local Features from RGB-D Data for One-Shot Learning Gesture Recognition.

Availability of handy RGB-D sensors has brought about a surge of gesture recognition research and applications. Among various approaches, one shot learning approach is advantageous because it requires minimum amount of data. Here, we provide a thorough review about one-shot learning gesture recognition from RGB-D data and propose a novel spatiotemporal feature extracted from RGB-D data, namely mixed features around sparse keypoints (MFSK). In the review, we analyze the challenges that we are facing, and point out some future research directions which may enlighten researchers in this field. The proposed MFSK feature is robust and invariant to scale, rotation and partial occlusions. To alleviate the insufficiency of one shot training samples, we augment the training samples by artificially synthesizing versions of various temporal scales, which is beneficial for coping with gestures performed at varying speed. We evaluate the proposed method on the Chalearn gesture dataset (CGD). The results show that our approach outperforms all currently published approaches on the challenging data of CGD, such as translated, scaled and occluded subsets. When applied to the RGB-D datasets that are not one-shot (e.g., the Cornell Activity Dataset-60 and MSR Daily Activity 3D dataset), the proposed feature also produces very promising results under leave-one-out cross validation or one-shot learning.

Auswirkungen fußballspezifischer Belastungen auf den Bewegungsapparat

Soccer as a Stop-and-Go-sport goes along with a high level of physical strain on the locomotor system. Compared to similar kinds of sports, soccer is characterized by a high prevalence of overloads/injuries in the pelvic region. Since soccer frequently involves one-sided shot-training, modifications in the pelvic statics are possible. In a pilot study including 15 football-players-FP (age 26.9 +/- 3.1 yrs; 4.4 +/- 0.4 training units/week+ 1 leaque game) the pelvic statics was measured using the 3-d-recording system CMS70 (Zebris, Germany) directly before and after a defined shot training. The positions of the right and the left posterior superior iliac spines - PSIS were compared. Additionally, the stiffness of selective muscles was analyzed. Before intervention the right PSIS was heightened in 8 FP. In 4 FP the pelvic statics was balanced and in 3 FP the right PSIS was abased. After the shot training, the pelvic statics was balanced in 4 FP. In contrast, 10 FP showed a higher right PSIS and 1 FP had a lower right PSIS. However, modifications of the pelvic statics were detected in all directions. Our study demonstrates modification of the pelvic statics by asymmetric soccer-specific strains, but the reactions were individually different. It is possible, that changes in the pelvic statics may lead to changed function or overstrain of advertising muscles.

POLYNOMIAL NETWORKS VERSUS OTHER TECHNIQUES IN TEXT CATEGORIZATION

Many techniques and algorithms for automatic text categorization had been devised and proposed in the literature. However, there is still much space for researchers in this area to improve existing algorithms or come up with new techniques for text categorization (TC). Polynomial Networks (PNs) were never used before in TC. This can be attributed to the huge datasets used in TC, as well as the technique itself which has high computational demands. In this paper, we investigate and propose using PNs in TC. The proposed PN classifier has achieved a competitive classification performance in our experiments. More importantly, this high performance is achieved in one shot training (noniteratively) and using just 0.25%–0.5% of the corpora features. Experiments are conducted on the two benchmark datasets in TC: Reuters-21578 and the 20 Newsgroups. Five well-known classifiers are experimented on the same data and feature subsets: the state-of-the-art Support Vector Machines (SVM), Logistic Regression (LR), the k-nearest-neighbor (kNN), Naive Bayes (NB), and the Radial Basis Function (RBF) networks.

Development of Laser Rifle Trainer with Full Shot Imitation

Laser trainers are effective tools for shot and sportsmen training. However, the majority of trainers have neither realistic recoil nor sound imitation systems. The objective of the development of laser rifle trainer with full shot imitation was to investigate and simulate the recoil of combat weapons under single and serial shooting regimes so that the training weapons could simulate complete recoil and sound imitation.Theoretical and experimental investigations lead to the development of shot training systems which were successfully implemented in Lithuanian Army combat training facilities. The laser trainers are also equipped with interactive software interface which does not only control and register the process of shooting but also can reproduce the whole set of statistical data, including psychological impact on the rifleman and detection of his training problems. The incorporation of multimedia advances the system up to interactive laser shot trainer in virtual combat reality.