Action recognition based on a human skeleton is an extremely challenging research problem. The temporal information contained in the human skeleton is more difficult to extract than the spatial information. Many researchers focus on graph convolution networks and apply them to action recognition. In this study, an action recognition method based on a two-stream network called RNXt-GCN is proposed on the basis of the Spatial-Temporal Graph Convolutional Network (ST-GCN). The human skeleton is converted first into a spatial-temporal graph and a SkeleMotion image which are input into ST-GCN and ResNeXt, respectively, for performing the spatial-temporal convolution. The convolved features are then fused. The proposed method models the temporal information in action from the amplitude and direction of the action and addresses the shortcomings of isolated temporal information in the ST-GCN. The experiments are comprehensively performed on the four datasets: 1) UTD-MHAD, 2) Northwestern-UCLA, 3) NTU RGB-D 60, and 4) NTU RGB-D 120. The proposed model shows very competitive results compared with other models in our experiments. On the experiments of NTU RGB?+?D 120 dataset, our proposed model outperforms those of the state-of-the-art two-stream models.
This study proposes a method for in situ local deoxyribonucleic acid (DNA) replication and detection in a long DNA strand through laser‐induced heating and strong avidin–biotin binding. To achieve the target DNA replication, dielectrophoresis was generated to stretch and immobilise DNA strands on both ends of the electrode. Subsequently, local DNA sequences were replicated using thermal cycles generated by laser‐induced heating. Replicated double‐stranded DNA products were captured in situ on a solid surface and detected using the fluorescence intensity of quantum dots (Qdots). The results revealed that after six laser‐induced thermal cycles, the replicated local DNA sequence could be detected by analysing the difference between Qdot fluorescent intensity before and after replication. The proposed method is expected to improve the efficiency of biosample gene sequence analysis.Inspec keywords: DNA, molecular biophysics, fluorescence, electrophoresis, biochemistry, molecular configurations, quantum dots, laser beam applications, biothermicsOther keywords: laser‐induced heating, long DNA strand, target DNA replication, DNA strands, local DNA sequences, thermal cycles, replicated double‐stranded DNA products, replicated local DNA sequence, in situ DNA replication, in situ local deoxyribonucleic acid replication, strong avidin‐biotin binding, biosample gene sequence analysis, Qdot fluorescent intensity, laser‐induced thermal cycles相似文献
PURPOSE: To review the current literature and generate recommendations on the role of newer technology in the management of the unanticipated difficult airway. METHODS: A literature search using key words and filters of English language and English abstracted publications from 1990-96 contained in the Medline, Current Contents and Biological Abstracts databases was carried out. The literature was reviewed and condensed and a series of evidence-based recommendations were evolved. CONCLUSIONS: The unanticipated difficult airway occurs with a low but consistent incidence in anaesthesia practice. Difficult direct laryngoscopy occurs in 1.5-8.5% of general anaesthetics and difficult intubation occurs with a similar incidence. Failed intubation occurs in 0.13-0.3% general anaesthetics. Current techniques for predicting difficulty with laryngoscopy and intubation are sensitive, non-specific and have a low positive predictive value. Assessment techniques which utilize multiple characteristics to derive a risk factor tend to be more accurate predictors. Devices such as the laryngeal mask, lighted stylet and rigid fibreoptic laryngoscopes, in the setting of unanticipated difficult airway, are effective in establishing a patient airway, may reduce morbidity and are occasionally lifesaving. Evidence supports their use in this setting as either alternatives to facemask and bag ventilation, when it is inadequate to support oxygenation, or to the direct laryngoscope, when tracheal intubation has failed. Specifically, the laryngeal mask and Combitube have proved to be effective in establishing and maintaining a patent airway in "cannot ventilate" situations. The lighted stylet and Bullard (rigid) fibreoptic scope are effective in many instances where the direct laryngoscope has failed to facilitate tracheal intubation. The data also support integration of these devices into strategies to manage difficult airway as the new standard of care. Training programmes should ensure graduate physicians are trained in the use of these alternatives. Continuing medical education courses should allow physicians in practice the opportunity to train with these alternative devices. 相似文献
Transmission Control Protocol (TCP) provides mechanisms for reliable data communications. Although it works well in wired networks, it fails to offer satisfactory performance in lossy and wireless environments. And in the multi-hop wireless infrastructure, packet delivery suffers high cumulative loss rate if traveling over multiple wireless hops. The Selective acknowledgment (SACK) is one header option that might be used to combat segment corruptions in air channels. In this paper, an alternative set of flow control mechanisms is proposed to handle high packet loss rate in a wireless medium. Using a measurement-based mechanism, sustainable segment delivery is achievable through a novel size-reduction method. Multiple segment retransmission mechanisms are introduced to reduce successive timeout events. One single byte loss is sufficient to waste all other bytes in a file received at a destination. That is, a good TCP flow control mechanism should provide a high successful file transmission completion rate, and this is set as our design goal. Through thorough simulations, our proposed multi-segment retransmission designs perform with higher successful file transfer rate and fewer timeout events than NewReno and SACK under a wide range of packet loss probabilities. 相似文献
Abstract— Owing to the fast developments of computer networks and color imaging technologies, there is a tendency towards remote proofing in the printing or textile industries. More cases such as cross‐display color matching will occur in the future information society. The problems of color mismatching we usually find in cross‐display devices have been widely discussed today. In this paper, to achieve equivalent color matching between a sRGB monitor and an Adobe RGB monitor, the spatial gamut mapping algorithm (SGMA), which introduces sigmoidal tone mapping, multi‐mapping paths, and unsharp mask (USM) operation into a sRGB color‐management system, is proposed. According to the designs of USM locations, this proposed SGMA can be further developed into pre‐USM, post‐USM, and double‐USM types. Besides, two critical image characteristics, edge map and color histogram, are investigated to establish the relationship between image content and SGMA. The psychophysical experimental results show that double‐USM SGMA obtains better color matching than the other spatial types. 相似文献
Why would people use Short Message Service (SMS) to say something they would not say in person? There is a trend that SMS is becoming more and more popular because it facilitates more extended modes of communications. Using technology acceptance model, we hypothesize the attitude of SMS would be influenced by its perceived effectiveness for communications, perceived ease of use, and subjective norm. Besides the special aspects of communications, conflict avoidance and privacy protection will enforce the impact of perceived effectiveness of SMS for communication. We investigated 953 SMS users and the results support most of our hypotheses. Furthermore, our analyses also show there are differences between females and males on the influence mechanism behind their attitude towards SMS.相似文献