Search

Abstract Human pose estimation is an important task in computer vision, which can provide key point detection of human body and obtain bone information. At present, human pose estimation is mainly utilized for detection of large targets, and there is no solution for detection of small targets. This paper proposes a multi-channel spatial information feature based human pose (MCSF-Pose) estimation algorithm to address the issue of medium and small targets inaccurate detection of human key points in scenarios involving occlusion and multiple poses. The MCSF-Pose network is a bottom-up regression network. Firstly, an UP-Focus module is designed to expand the feature information while reducing parameter computation during the up-sampling process. Then, the channel segmentation strategy is adopted to cut the features, and the feature information of multiple dimensions is retained through different convolutional groups, which reduces the parameter lightweight network model and makes up for the loss of the feature information associated with the depth of the network. Finally, the three-layer PANet structure is designed to reduce the complexity of the model. With the aid of the structure, it also to improve the detection accuracy and anti-interference ability of human key points. The experimental results indicate that the proposed algorithm outperforms YOLO-Pose and other human pose estimation algorithms on COCO2017 and MPII human pose datasets.

Human pose estimation is an important task in computer vision, which can provide key point detection of human body and obtain bone information. At present, human pose estimation is mainly utilized for detection of large targets, and there is no solution for detection of small targets. This paper proposes a multi-channel spatial information feature based human pose (MCSF-Pose) estimation algorithm to address the issue of medium and small targets inaccurate detection of human key points in scenarios involving occlusion and multiple poses. The MCSF-Pose network is a bottom-up regression network. Firstly, an UP-Focus module is designed to expand the feature information while reducing parameter computation during the up-sampling process. Then, the channel segmentation strategy is adopted to cut the features, and the feature information of multiple dimensions is retained through different convolutional groups, which reduces the parameter lightweight network model and makes up for the loss of the feature information associated with the depth of the network. Finally, the three-layer PANet structure is designed to reduce the complexity of the model. With the aid of the structure, it also to improve the detection accuracy and anti-interference ability of human key points. The experimental results indicate that the proposed algorithm outperforms YOLO-Pose and other human pose estimation algorithms on COCO2017 and MPII human pose datasets. [ABSTRACT FROM AUTHOR]

This article presents research conducted on the backplane high‐speed Peripheral Component Interconnect Express version 4 (PCIe gen4) link between central processor root complex and the end‐point device in order to confirm its compliance. Signal integrity analysis was performed using a method that included channel segmentation, development of corner models for channel segments, full‐wave electromagnetic simulation utilizing the finite element method (FEM) and the creation of cascaded channel models, as well as determination of the worst‐case models for the channel. The results of this analysis were then used to modify the channel design in order to decrease insertion loss. Changes in the design included utilizing a very low loss dielectric for the backplane board, correcting the sizes of differential pairs and decreasing the via stubs. Frequency domain simulation of the modified channel design has shown a 25% decrease in insertion loss. The calculation of eye diagrams for the modified design has confirmed that the channel parameters meet the requirements for the PCIe gen4 standard. [ABSTRACT FROM AUTHOR]

Abstract This article presents research conducted on the backplane high‐speed Peripheral Component Interconnect Express version 4 (PCIe gen4) link between central processor root complex and the end‐point device in order to confirm its compliance. Signal integrity analysis was performed using a method that included channel segmentation, development of corner models for channel segments, full‐wave electromagnetic simulation utilizing the finite element method (FEM) and the creation of cascaded channel models, as well as determination of the worst‐case models for the channel. The results of this analysis were then used to modify the channel design in order to decrease insertion loss. Changes in the design included utilizing a very low loss dielectric for the backplane board, correcting the sizes of differential pairs and decreasing the via stubs. Frequency domain simulation of the modified channel design has shown a 25% decrease in insertion loss. The calculation of eye diagrams for the modified design has confirmed that the channel parameters meet the requirements for the PCIe gen4 standard.

We describe a compact terahertz (THz) SPICE/advanced design system (ADS) model based on the extended Enz-Krummenacher-Vittoz (EKV) MOSFET model with channel segmentation but accounting for the new physics of the electron transport at THz frequencies. The model includes the distributed Drude inductance representing the electron inertia. The channel segmentation allows reproducing the nonuniformity in the electron density excitations and the plasmonic oscillations at THz frequencies. The model was validated by comparing the simulation results with the analytical THz detection theory and experimental data for silicon MOSFETs and FinFETs with gate lengths ranging from 60 to 800 nm. The good agreement between the simulated and analytical results and experimental data revealed the significance of the electron inertia and demonstrated the improved validity of the model at higher frequencies with an increase in the number of segments. The model was used to simulate a novel TeraFET THz spectrometer and the resonant TeraFET THz detector. These results show the model utility for device and circuit simulations at sub-THz and THz frequencies. An important application of the THz spectrometer simulation is in the design of a single field-effect transistor (FET) THz vector detector that could be extremely useful for 5G (and beyond) line-of-sight applications. [ABSTRACT FROM AUTHOR]

Background: Ethnic minority women from low-income countries who live in high-income countries are more physically inactive than ethnic majority women in those countries. At the same time, they can be harder to reach with health promotion programs. Targeting recruitment channels and execution to ethnic groups could increase reach and receptivity to program participation. We explored using ethnically specific channels and key figures to reach Ghanaian, Antillean, and Surinamese mothers with an invitation for an exercise program, and subsequently, to determine the mothers' receptivity and participation. Methods: We conducted a mixed methods process evaluation in Amsterdam, the Netherlands. To recruit mothers, we employed ethnically specific community organizations and ethnically matched key figures as recruiters over Dutch health educators. Reach and participation were measured using reply cards and the attendance records from the exercise programs. Observations were made of the recruitment process. We interviewed 14 key figures and 32 mothers to respond to the recruitment channel and recruiter used. Content analysis was used to analyze qualitative data. Results: Recruitment through ethnically specific community channels was successful among Ghanaian mothers, but less so among Antillean and Surinamese mothers. The more close-knit an ethnic community was, retaining their own culture and having poorer comprehension of the Dutch language, the more likely we were to reach mothers through ethnically specific organizations. Furthermore, we found that using ethnically matched recruiters resulted in higher receptivity to the program and, among the Ghanaian mothers in particular, in greater participation. This was because the ethnically matched recruiter was a familiar, trusted person, a translator, and a motivator who was enthusiastic, encouraging, and able to adapt her message (targeting/tailoring). Using a health expert was preferred in order to increase the credibility and professionalism of the recruitment. Conclusions: Recruitment for an exercise program through ethnically specific organizations seems to contribute to its reach, particularly in close-knit, highly organized ethnic communities with limited fluency in the local language. Using ethnically matched recruiters as motivator, translator, and trusted person seems to enhance receptivity of a health promotion program. An expert is likely to be needed for effective information delivery. [ABSTRACT FROM AUTHOR]

As a challenging computer vision task, video-based person Re-IDentification (Re-ID) has been intensively studied, and recent works have achieved a series of satisfactory results by capturing spatial temporal relationships. However, extensive observations have found that the same feature vector generated by a convolutional neural network contains considerable redundant information in the channel dimension. This issue is seldom investigated. A Spatial Temporal and Channel Aware Network (STCAN) for video-based ReID is studied in this paper. It jointly considers spatial temporal and channel information. Firstly, the Spatial Attention Enhanced (SAE) convolutional network is developed as the backbone network to learn spatial enhanced features from video frames. Secondly, a Channel Segmentation and Group Shuffle (CSGS) convolution module is designed to jointly address temporal and channel relations. Finally, a Two Branch Weighted Fusion (TBWF) mechanism is introduced to enhance the robustness of the Re-ID network by fusing the output of the SAE backbone network and CSGS. Comprehensive experiments are conducted on three large-scale datasets MARS, LSVID, and P-DESTRE. The experimental results imply that the STCAN can effectively improve the performance of video-based Re-ID and outperform several state-of-the-art methods. • A feature aggregation module, CSGS, is proposed to capture temporal and channel relations. • An attention mechanism, SAE, is introduced to enhance spatial dependencies. • A Fusion mechanism is proposed to combine the output of SAE and CSGS [ABSTRACT FROM AUTHOR]

Abstract: This paper presents an efficient formulation of a channel segmentation based approach to non-quasi-static modelling of the MOS transistor, in the context of a charge-based MOSFET model. In this minimal channel segmentation approach, only the essential charge equations are evaluated for each channel segment while other effects are handled at device level. As a result, simulation time is drastically reduced compared to a full channel segmentation approach. The model is validated versus measurement up to 10GHz and passes relevant benchmark tests. [Copyright &y& Elsevier]