scholarly journals Semantic-Segmentation-Based Rail Fastener State Recognition Algorithm

2021 ◽  
Vol 2021 ◽  
pp. 1-15
Author(s):  
Liming Li ◽  
Rui Sun ◽  
Shuguang Zhao ◽  
Xiaodong Chai ◽  
Shubin Zheng ◽  
...  
Keyword(s):  
Semantic Segmentation ◽  
Recognition Algorithm ◽  
Detection Methods ◽  
State Recognition ◽  
Detection Model ◽  
Practical Engineering ◽  
The One ◽  
And Function ◽  
Key Steps ◽  
Engineering Projects

Rail fastener status recognition and detection are key steps in the inspection of the rail area status and function of real engineering projects. With the development of and widespread interest in image processing techniques and deep learning theory, detection methods that combine the two have yielded promising results in practical detection applications. In this paper, a semantic-segmentation-based algorithm for the state recognition of rail fasteners is proposed. On the one hand, we propose a functional area location and annotation method based on a salient detection model and construct a novel slab-fastclip-type rail fastener dataset. On the other hand, we propose a semantic-segmentation-framework-based model for rail fastener detection, where we detect and classify rail fastener states by combining the pyramid scene analysis network (PSPNet) and vector geometry measurements. Experimental results prove the validity and superiority of the proposed method, which can be introduced into practical engineering projects.

Recognition of classroom student state features based on deep learning algorithms and machine learning

2020 ◽  
pp. 1-12
Author(s):  
Hu Jingchao ◽  
Haiying Zhang
Keyword(s):  
Deep Learning ◽  
Feature Recognition ◽  
Subjective Evaluation ◽  
Recognition Algorithm ◽  
Video Data ◽  
State Recognition ◽  
Detection Model ◽  
State Classification ◽  
Intelligent Models ◽  
Model Recognition

The difficulty in class student state recognition is how to make feature judgments based on student facial expressions and movement state. At present, some intelligent models are not accurate in class student state recognition. In order to improve the model recognition effect, this study builds a two-level state detection framework based on deep learning and HMM feature recognition algorithm, and expands it as a multi-level detection model through a reasonable state classification method. In addition, this study selects continuous HMM or deep learning to reflect the dynamic generation characteristics of fatigue, and designs random human fatigue recognition experiments to complete the collection and preprocessing of EEG data, facial video data, and subjective evaluation data of classroom students. In addition to this, this study discretizes the feature indicators and builds a student state recognition model. Finally, the performance of the algorithm proposed in this paper is analyzed through experiments. The research results show that the algorithm proposed in this paper has certain advantages over the traditional algorithm in the recognition of classroom student state features.

scholarly journals An Efficient Pedestrian Detection Method Based on YOLOv2

2018 ◽  
Vol 2018 ◽  
pp. 1-10 ◽  
Author(s):  
Zhongmin Liu ◽  
Zhicai Chen ◽  
Zhanming Li ◽  
Wenjin Hu
Keyword(s):  
Detection Method ◽  
Size Range ◽  
State Of The Art ◽  
Pedestrian Detection ◽  
Semantic Segmentation ◽  
Detection Methods ◽  
Detection Model ◽  
Great Progress ◽  
Good Trade ◽  
Speed And Accuracy

In recent years, techniques based on the deep detection model have achieved overwhelming improvements in the accuracy of detection, which makes them being the most adapted for the applications, such as pedestrian detection. However, speed and accuracy are a pair of contradictions that always exist and have long puzzled researchers. How to achieve the good trade-off between them is a problem we must consider while designing the detectors. To this end, we employ the general detector YOLOv2, a state-of-the-art method in the general detection tasks, in the pedestrian detection. Then we modify the network parameters and structures, according to the characteristics of the pedestrians, making this method more suitable for detecting pedestrians. Experimental results in INRIA pedestrian detection dataset show that it has a fairly high detection speed with a small precision gap compared with the state-of-the-art pedestrian detection methods. Furthermore, we add weak semantic segmentation networks after shared convolution layers to illuminate pedestrians and employ a scale-aware structure in our model according to the characteristics of the wide size range in Caltech pedestrian detection dataset, which make great progress under the original improvement.

scholarly journals Driver Fatigue Detection Based on Facial Key Points and LSTM

2021 ◽  
Vol 2021 ◽  
pp. 1-9
Author(s):  
Long Chen ◽  
Guojiang Xin ◽  
Yuling Liu ◽  
Junwei Huang
Keyword(s):  
Traffic Safety ◽  
Short Term Memory ◽  
Field Research ◽  
Recognition Algorithm ◽  
Detection Methods ◽  
Detection Accuracy ◽  
State Recognition ◽  
Key Points ◽  
Fatigue Driving ◽  
Fatigue Detection

In recent years, fatigue driving has been a serious threat to the traffic safety, which makes the research of fatigue detection a hotspot field. Research on fatigue recognition has a great significance to improve the traffic safety. However, the existing fatigue detection methods still have room for improvement in detection accuracy and efficiency. In order to detect whether the driver has fatigue driving, this paper proposes a fatigue state recognition algorithm. The method first uses MTCNN (multitask convolutional neural network) to detect human face, and then DLIB (an open-source software library) is used to locate facial key points to extract the fatigue feature vector of each frame. The fatigue feature vectors of multiple frames are spliced into a temporal feature sequence and sent to the LSTM (long short-term memory) network to obtain a final fatigue feature value. Experiments show that compared with other methods, the fatigue state recognition algorithm proposed in this paper has achieved better results in accuracy. The average accuracy of the proposed method in detecting key points of the face is as high as 93%, and the running time is less than half of the ordinary DLIB method.

Comparative study on e-bikes’ decision-making behaviors under flashing green and green countdown

2021 ◽  
pp. 1-8
Author(s):  
Haoyang Meng ◽  
Sheng Dong ◽  
Jibiao Zhou ◽  
Shuichao Zhang ◽  
Zhenjiang Li
Keyword(s):  
Decision Making ◽  
Critical Issue ◽  
Trajectory Data ◽  
Decision Points ◽  
Binary Logit ◽  
The One ◽  
Common Signal ◽  
And Function ◽  
Decision Behaviors ◽  
Stop Line

Green flash light (FG) and green countdown (GC) are the two most common signal formats applied in green-red transition that provides drivers additional alert before termination of green phase. Due to their importance and function in stop-pass decision-making process, proper use of them has become a critical issue to greatly improve the safety and efficiency of signalized intersections. Gradually e-bike riders have become more important commuters in China, however, the influence of FG or GC on them is not clear yet and need pay more attention to it. This study chooses two almost identical intersections to obtain highly accurate trajectory data of e-bike riders to study their decision-making behaviors under FG or GC. The e-bike riders’ behavior is classified into four categories and is to identify their stop-pass decision points using the acceleration trend. Two binary-logit models were built to predict the stop–pass decision behaviors for the different e-bike rider groups, explaining that the potential time to the stop-line is the dominant independent factor of the different behaviors of GC and FG. Furthermore empirical analysis of decision points indicated that GC provides the earlier stop-pass decision point and longer decision making duration on the one side while results in more complexity of decision making and greater risk of stop-line crossing than FG on the other side.

scholarly journals Multiview deep learning based on tensor decomposition and its application in fault detection of overhead contact systems

2021 ◽  
Author(s):  
Xuewu Zhang ◽  
Yansheng Gong ◽  
Chen Qiao ◽  
Wenfeng Jing
Keyword(s):  
High Speed ◽  
Tensor Decomposition ◽  
Detection Methods ◽  
Detection Accuracy ◽  
Feature Maps ◽  
Training Time ◽  
Detection Model ◽  
Railway Line ◽  
Result Show ◽  
Deep Layers

AbstractThis article mainly focuses on the most common types of high-speed railways malfunctions in overhead contact systems, namely, unstressed droppers, foreign-body invasions, and pole number-plate malfunctions, to establish a deep-network detection model. By fusing the feature maps of the shallow and deep layers in the pretraining network, global and local features of the malfunction area are combined to enhance the network's ability of identifying small objects. Further, in order to share the fully connected layers of the pretraining network and reduce the complexity of the model, Tucker tensor decomposition is used to extract features from the fused-feature map. The operation greatly reduces training time. Through the detection of images collected on the Lanxin railway line, experiments result show that the proposed multiview Faster R-CNN based on tensor decomposition had lower miss probability and higher detection accuracy for the three types faults. Compared with object-detection methods YOLOv3, SSD, and the original Faster R-CNN, the average miss probability of the improved Faster R-CNN model in this paper is decreased by 37.83%, 51.27%, and 43.79%, respectively, and average detection accuracy is increased by 3.6%, 9.75%, and 5.9%, respectively.

scholarly journals CNN-Based Road-Surface Crack Detection Model That Responds to Brightness Changes

Electronics ◽  
2021 ◽  
Vol 10 (12) ◽  
pp. 1402
Author(s):  
Taehee Lee ◽  
Yeohwan Yoon ◽  
Chanjun Chun ◽  
Seungki Ryu
Keyword(s):  
Surface Crack ◽  
Autonomous Vehicles ◽  
Crack Detection ◽  
Model Performance ◽  
Semantic Segmentation ◽  
Road Surface ◽  
Image Brightness ◽  
The Road ◽  
Detection Model ◽  
Artificial Intelligence Models

Poor road-surface conditions pose a significant safety risk to vehicle operation, especially in the case of autonomous vehicles. Hence, maintenance of road surfaces will become even more important in the future. With the development of deep learning-based computer image processing technology, artificial intelligence models that evaluate road conditions are being actively researched. However, as the lighting conditions of the road surface vary depending on the weather, the model performance may degrade for an image whose brightness falls outside the range of the learned image, even for the same road. In this study, a semantic segmentation model with an autoencoder structure was developed for detecting road surface along with a CNN-based image preprocessing model. This setup ensures better road-surface crack detection by adjusting the image brightness before it is input into the road-crack detection model. When the preprocessing model was applied, the road-crack segmentation model exhibited consistent performance even under varying brightness values.

Structural Characterization of Complex Bacterial Glycolipids by Fourier Transform Mass Spectrometry

2005 ◽  
Vol 11 (5) ◽  
pp. 535-546 ◽  
Author(s):  
Anna Kondakov ◽  
Buko Lindner
Keyword(s):  
Mass Spectrometry ◽  
Fourier Transform ◽  
Structural Information ◽  
Adaptive Immune System ◽  
Ion Cyclotron Resonance ◽  
Bacterial Membranes ◽  
Multiphoton Dissociation ◽  
The One ◽  
And Function

Bacterial glycolipids are complex amphiphilic molecules which are, on the one hand, of utmost importance for the organization and function of bacterial membranes and which, on the other hand, play a major role in the activation of cells of the innate and adaptive immune system of the host. Already small alterations to their chemical structure may influence the biological activity tremendously. Due to their intrinsic biological heterogeneity [number and type of fatty acids, saccharide structures and substitution with for example, phosphate ( P), 2-aminoethyl-(pyro)phosphate groups ( P-Etn) or 4-amino-4-deoxyarabinose (Ara4N)], separation of the different components are a prerequisite for unequivocal chemical and nuclear magnetic resonance structural analyses. In this contribution, the structural information which can be obtained from heterogenous samples of glycolipids by Fourier transform (FT) ion cyclotron resonance mass spectrometric methods is described. By means of recently analysed complex biological samples, the possibilities of high-resolution electrospray ionization FT-MS are demonstrated. Capillary skimmer dissociation, as well as tandem mass spectrometry (MS/MS) analysis utilizing collision-induced dissociation and infrared multiphoton dissociation, are compared and their advantages in providing structural information of diagnostic importance are discussed.

Socialisation Tactics of the Spectacled Parrotlet (Forpus Conspicillatus)

Behaviour ◽  
1991 ◽  
Vol 119 (1-2) ◽  
pp. 1-29 ◽  
Author(s):  
Kyra Garnetzke-Stollmann ◽  
Dierk Franck
Keyword(s):  
Sexual Behaviour ◽  
Early Stage ◽  
Rank Order ◽  
Host Group ◽  
Pair Bonds ◽  
Group Members ◽  
The One ◽  
And Function ◽  
Sibling Group ◽  
Sibling Groups

AbstractSpectacled parrotlets live in a complex system of individual relationships throughout their lives. The adults form exclusive pair bonds, addressing all friendly and sexual behaviour patterns to each other. Pair mates cooperate in agonistic situations. As long as they stay close together they hold the same rank-order position. In mate-choice experiments females (not males) significantly preferred a mate which formerly held a high social position. There are also non-exclusive pair bonds, which are far less stable than exclusive ones. Only exclusive pairs have a good chance to occupy a breeding cavity. All group members are synchronized in many of their activities, such as foraging, preening or resting. They are keenly interested in unusual activities of other group members. Social learning, including copying sexual techniques, seems to be essential. After fledging the parents keep their offspring at a distance from a very early stage. Instead of a close parent-offspring relationship the fledglings form sibling groups with their nest mates. Over a period of months siblings remain the main interaction partners for all friendly and playful activities. They also support one another in agonistic situations. In the first months of life even courtship feeding and sexual behaviour are addressed predominantly to siblings. Thus a pair-like relationship is established between siblings, anticipating the permanent pair bond of adults. Single fledglings, deprived of the experience of a sibling group, remained poorly integrated into the group. They developed alternative socialisation tactics, namely (1) joining a host group of unrelated siblings, (2) renewing a friendly partnership with the parents, (3) helping to protect and feed younger siblings or even unrelated fledglings and (4) seeking early partnership with unrelated group members. Out of 10 single fledglings only the one that was accepted by a host sibling group immediately after fledging became well integrated into the whole group and reproduced well. It is argued that sibling groups offer good opportunities for learning partnership and function as a safe basis for exploring the social environment. It is tentatively proposed that single fledglings have a decreased probability of reproductive success.

Design of Gait Recognition System Based on FPGA and DSP

2014 ◽  
Vol 687-691 ◽  
pp. 3861-3868
Author(s):  
Zheng Hong Deng ◽  
Li Tao Jiao ◽  
Li Yan Liu ◽  
Shan Shan Zhao
Keyword(s):  
Monitoring System ◽  
Nearest Neighbor ◽  
Gait Recognition ◽  
Recognition Rate ◽  
Digital Signal ◽  
Recognition System ◽  
Recognition Algorithm ◽  
Zernike Moment ◽  
Intelligent Monitoring ◽  
The One

According to the trend of the intelligent monitoring system, on the basis of the study of gait recognition algorithm, the intelligent monitoring system is designed based on FPGA and DSP; On the one hand, FPGA’s flexibility and fast parallel processing algorithms when designing can be both used to avoid that circuit can not be modified after designed; On the other hand, the advantage of processing the digital signal of DSP is fully taken. In the feature extraction and recognition, Zernike moment is selected, at the same time the system uses the nearest neighbor classification method which is more mature and has good real-time performance. Experiments show that the system has high recognition rate.

L'etica nella professione forense

2009 ◽  
pp. 163-172
Author(s):  
Angelo Abignente
Keyword(s):  
Public Interests ◽  
Legal Proceedings ◽  
Legal Professionals ◽  
Positive Law ◽  
Conflicting Interests ◽  
Discursive Participation ◽  
The Moment ◽  
The One ◽  
And Function ◽  
Role And Function

- The positive law tradition has hitherto had nothing to say about the legal profession's role and function, focusing more interest on questions of justice, of the legitimisation of power and of the genesis and organisation of normative material. This trend is now subject to a reversal promoted by new, neo-constitutionalist, narrativist, analytical and hermeneutic experiences, which no longer focuses attention on the moment when law is produced, but on the one when it is applied, reappraising and revitalising the function of the judge, of the attorneys and of other legal professionals. The attorney becomes an active protagonist, an intermediary not only between conflicting interests in a controversy, but also between opposing public interests, while the reappraisal of his role stimulates thinking about the ethical dimension of how the legal profession is practised. Referring to the theories of Habermas and of Alexy, the author treats the reasonable status of argumentation as the supreme ethical instance necessary for a decision that interferes in the sphere of another person's action. At the same time, however, the control of the reasonable status of the respective arguments on both sides is the ethical instance required of the attorneys taking part in the legal proceedings. It takes the form of compliance with the rules characteristic of the practical discourse, primarily the rule of free discursive participation that enables the onus of the argumentation to be explained. Ernesto de

Sign in / Sign up

Export Citation Format

Share Document