Virtual try-on system under arbitrary human poses has huge application potential, yet raises quite a lot of challenges, e.g. self-occlusions, heavy misalignment among diverse poses, and diverse clothes textures. Existing methods aim at fitting new clothes into a person can only transfer clothes on the fixed human pose, but still show unsatisfactory performances which often fail to preserve the identity, lose the texture details, and decrease the diversity of poses. In this paper, we make the first attempt towards multi-pose guided virtual try-on system, which enables transfer clothes on a person image under diverse poses. Given an input person image, a desired clothes image, and a desired pose, the proposed Multi-pose Guided Virtual Try-on Network (MG-VTON) can generate a new person image after fitting the desired clothes into the input image and manipulating human poses. Our MG-VTON is constructed in three stages: 1) a desired human parsing map of the target image is synthesized to match both the desired pose and the desired clothes shape; 2) a deep Warping Generative Adversarial Network (Warp-GAN) warps the desired clothes appearance into the synthesized human parsing map and alleviates the misalignment problem between the input human pose and desired human pose; 3) a refinement render utilizing multi-pose composition masks recovers the texture details of clothes and removes some artifacts. Extensive experiments on well-known datasets and our newly collected largest virtual try-on benchmark demonstrate that our MG-VTON significantly outperforms all state-of-the-art methods both qualitatively and quantitatively with promising multipose virtual try-on performances.

No abstract

No abstract

Purpose This paper aims to examine the different impacts of six variables on firm technological innovation performance in different high-tech industries in China. Through a comparative analysis of data about growth enterprises market board (GEM)-listed companies, this study attempts to get some conclusions, to help firms in different high-tech industries use resources more rationally and to improve technological innovation performance more effectively. Design/methodology/approach This paper constructs semi-parametric models based on the relevant data of GEM-listed companies during 2010 to 2015 for different high-tech industries. These models can ensure that the influencing factors of firm technological innovation performance are no longer restricted to a particular aspect but can provide a comprehensive comparative analysis of the effects of factors on firm technological innovation performance in different high-tech industries. Findings The empirical results show that R&D expenditures have a significant positive impact on firm technological innovation performance in most high-tech industries, but not in electronic and communication equipment manufacturing industry; R&D personnel investment and government subsidies have significant positive impacts on firm technological innovation performance in knowledge-oriented industries; technology diversity has a significant positive impact on firm technological innovation performance in technology-oriented industries; the proportion of exports shows an inverted U-shaped relationship with firm technological innovation performance in electronic and communication equipment manufacturing industry, while firm size shows an inverted U-shaped relationship with firm technological innovation performance in general equipment manufacturing industry; and the effect of semi-parametric model fit is superior to the general parameters model. Originality/value Drawing on the resource dependence perspective, this paper is the first to consider a comprehensive treatment of differential effects of internal resources (R&D personnel, R&D expenditure), external resources (government subsides) and firm characteristics (firm size, export ratio) on firm technological innovation performance in different high-tech industries in an emerging country, in particular in contrast to previous studies that have focused on a single industry or taken the type of industry as a control variable. In addition, most studies about the determinants of firm innovation performance are based on survey questionnaires, which may introduce large subjective errors. Setting the relationship between variables in advance may also introduce fit error when using a general-parameter model. Semi-parametric regression which is used in this paper is able to prevent this shortcoming effectively. When constructing a regression model, this can be exempted from the formal constraints, thus estimating data more accurately and ensuring superior fit.

A structure based on plasmonic nano-capillary resonators for optical wavelengths demultiplexing is proposed and numerically investigated. The structure consists of main/bus waveguide connected with series of nano-capillary resonators, each of which tuned at different wavelength transmission band. A model based on resonator theory is given to design the working wavelength of the structure. Both analytical and simulation results reveal that the demultiplexing wavelength of each channel has linear and nonlinear relationships with length and width of the nano-capillary structure.