MSSTNet:Multi-scale facial videos pulse extraction network based on separable spatiotemporal convolution and dimension separable attention

Background The use of remote photoplethysmography(rPPG)to estimate blood volume pulse in a noncontact manner has been an active research topic in recent years.Existing methods are primarily based on a singlescale region of interest(ROI).However,some noise signals that are not easily separated in a single-scale space can be easily separated in a multi-scale space.Also,existing spatiotemporal networks mainly focus on local spatiotemporal information and do not emphasize temporal information,which is crucial in pulse extraction problems,resulting in insufficient spatiotemporal feature modelling.Methods Here,we propose a multi-scale facial video pulse extraction network based on separable spatiotemporal convolution(SSTC)and dimension separable attention(DSAT).First,to solve the problem of a single-scale ROI,we constructed a multi-scale feature space for initial signal separation.Second,SSTC and DSAT were designed for efficient spatiotemporal correlation modeling,which increased the information interaction between the long-span time and space dimensions;this placed more emphasis on temporal features.Results The signal-to-noise ratio(SNR)of the proposed network reached 9.58dB on the PURE dataset and 6.77dB on the UBFC-rPPG dataset,outperforming state-of-the-art algorithms.Conclusions The results showed that fusing multi-scale signals yielded better results than methods based on only single-scale signals.The proposed SSTC and dimension-separable attention mechanism will contribute to more accurate pulse signal extraction.

Low band gap frequencies and multiplexing properties in 1D and 2D mass spring structures

This study reports on the propagation of elastic waves in 1D and 2D mass spring structures.An analytical and computation model is presented for the 1D and 2D mass spring systems with different examples.An enhancement in the band gap values was obtained by modeling the structures to obtain low frequency band gaps at small dimensions.Additionally,the evolution of the band gap as a function of mass value is discussed.Special attention is devoted to the local resonance property in frequency ranges within the gaps in the band structure for the corresponding infinite periodic lattice in the 1D and 2D mass spring system.A linear defect formed of a row of specific masses produces an elastic waveguide that transmits at the narrow pass band frequency.The frequency of the waveguides can be selected by adjusting the mass and stiffness coefficients of the materials constituting the waveguide.Moreover,we pay more attention to analyze the wave multiplexer and DE-multiplexer in the 2D mass spring system.We show that two of these tunable waveguides with alternating materials can be employed to filter and separate specific frequencies from a broad band input signal.The presented simulation data is validated through comparison with the published research,and can be extended in the development of resonators and MEMS verification.

Three Dimensional Separation with Spiral-Focus in a Decelerating Duct Flow (Effect of Asymmetric Inlet Boundary Layer Thickness)

An experimental apparatus was developed to study the three dimensional separated flow with spiral-foci. The internal decelerating flow was generated by the air suction from a side wall to produce the separation on an opposite-side wall. The relation between the upstream boundary layer and the generation of spiral-foci in the separation region was observed by a tuft method. As a result, it was clarified that the spiral-focus type separation could be produced on the side wall and its behavior was closely related to the vortices supplied into the separation region from the boundary layer developing along top wall or bottom one.