Person image generation through graph-based and appearance-decomposed generative adversarial network

Due to the sophisticated entanglements for non-rigid deformation, generating person images from source pose to target pose is a challenging work. In this paper, we present a novel framework to generate person images with shape consistency and appearance consistency. The proposed framework leverages the graph network to infer the global relationship of source pose and target pose in a graph for better pose transfer. Moreover, we decompose the source image into different attributes (e.g., hair, clothes, pants and shoes) and combine them with the pose coding through operation method to generate a more realistic person image. We adopt an alternate updating strategy to promote mutual guidance between pose modules and appearance modules for better person image quality. Qualitative and quantitative experiments were carried out on the DeepFashion dateset. The efficacy of the presented framework are verified.

Study on Automatic Tracking Method of Marking Points in Sports Image Sequence

: The effect of teaching and training in physical sports is improved by the sport demonstration system. The two-dimensional sport demonstration system is widely applied for the training of practical athletic. In the accurate motion positioning, a certain visual deficiency exists, and the two-dimensional sport demonstration system is analyzed by the kinematic. Methodology: Aiming at the problems in the real-time tracking of fast moving targets in sports images, an automatic tracking method of sports images, based on the registration of landmark points of the passive optical motion capture system, is proposed. First, build a human model and divide the human model into several limb segments. Then, find a corresponding relationship between the first frame of motion data and template data to complete the first frame of motion data registration; based on the smallest non-rigid deformation and point, set matching error to find the corresponding relationship between the current frame of motion data and the previous frame of registered motion data. Finally, through the mark points, follow up to complete the registration process of the marker. Results: Experiments show that the average processing accuracy of this algorithm can reach over 85%, and the processing time of a single frame of motion data is t<1/60s, which can meet real-time requirements. Conclusion: The multi-point set least squares matching algorithm is used to correct the registered landmark data rigidly. No manual intervention is required for the entire mark registration process.

External Dynamic Apparatus Correction in Rigid Posttraumatic Spinal De-formations Engineering and Clinical Aspects

Study design: The authors define most dangerous sections and calculate the stresses arising in the parts of the external fixation device. They configure external fixation devices taking into account the obtained results and use them in medical practice. Objective: the development of an apparatus correction technique for the treatment of spinal deformations. Summary of background data: standard methods of rigid deformation correction are technically complicated and high-costly; there are also possible formidable intraoperative complications. Authors consider a variant of apparatus external correction as an alternative to the known methods, which is devoid of the aforementioned disadvantages. Methods: The authors use theoretical and experimental research methods. The theoretical deductions are confirmed experimentally during clinical trials. Clinical studies were evaluated using ray-based diagnostic procedures, computed tomography, and patient interviewing using the Oswestry Disability Index and visual analogue scale. Results: The calculations showed that in case of one-sided point load, the stresses arising in the dangerous section of screws-rods from that load can reach 60 – 70% of the total stresses in the device nodes. During clinical testing it was established that external apparatus correction leads to normalization of anatomical relationships in the area of damaged segments, normalization of the spinal column axis, and reduction of pain; it does not preclude the use of dorsal internal puncture fixation. Conclusions: It is advisable to load the structural element of the external spinal fixation device in a skew-symmetric pattern, applying distraction and compression forces to the transverse plate on both sides along parallel straight lines in opposite directions. Apparatus external fixation makes it possible to perform a complete correction of rigid deformations without releasing the supporting structures, to control the level of neurological dis-orders, to simplify the technique of internal fixation operations, to use the minimally invasive procedure for the installation of the internal pedicle fixator.

Measuring change at Earth’s surface: On-demand vertical and three- dimensional topographic differencing implemented in OpenTopography

Topographic differencing measures landscape change by comparing multitemporal high-resolution topography data sets. Here, we focused on two types of topographic differencing: (1) Vertical differencing is the subtraction of digital elevation models (DEMs) that span an event of interest. (2) Three-dimensional (3-D) differencing measures surface change by registering point clouds with a rigid deformation. We recently released topo­graphic differencing in OpenTopography where users perform on-demand vertical and 3-D differencing via an online interface. OpenTopography is a U.S. National Science Foundation–funded facility that provides access to topographic data and processing tools. While topographic differencing has been applied in numerous research studies, the lack of standardization, particularly of 3-D differencing, requires the customization of processing for individ­ual data sets and hinders the community’s ability to efficiently perform differencing on the growing archive of topography data. Our paper focuses on streamlined techniques with which to efficiently difference data sets with varying spatial resolution and sensor type (i.e., optical vs. light detection and ranging [lidar]) and over variable landscapes. To optimize on-demand differencing, we considered algorithm choice and displacement resolution. The optimal resolution is controlled by point density, landscape characteristics (e.g., leaf-on vs. leaf-off), and data set quality. We provide processing options derived from metadata that allow users to produce optimal high-quality results, while experienced users can fine tune the parameters to suit their needs. We anticipate that the differencing tool will expand access to this state-of-the-art technology, will be a valuable educational tool, and will serve as a template for differencing the growing number of multitemporal topography data sets.

3-D dynamic simulation of knitwear based on the hybrid model

Due to the effect of force and deformation of fancy stitches, the simulation of knitwear is facing a great challenges of the real-time. In this paper, the physical-geometric hybrid method is applied to reduce the amount of calculation during the simulation of knitwear. Discretized Newton’s Method is used to analyze the gap between dynamic knitwear-human body and the knitwear model, and then the knitwear model is further divided into different regions. A three-dimensional (3-D) mesh for knitwear is constructed by the application of adaptive remeshing. This makes it possible to refine the mesh at the parts that need the presentation of fabric surface details. Simultaneously, it can merge the adjacent patches at parts without the requirement of showing the details, and form a large 3-D patch. In the light of regional division, the 3-D knitwear model is divided into the tight layer, floating layer, and loose layer. In addition, the geometric loop model based on the cuboid particle system is employed to simulate the real force of loops and knitwear for tight layer and loose layer. Near-rigid deformation method is also applied into the floating layer to improve simulation efficiency. In conclusion, the corresponding processing method is performed with different computational models, which brings the dynamic simulation effect of knitwear with realistic and real-time.

Calculation of temperature distribution and rheological properties of the lithosphere along transect IV in the Western Carpathian-Pannonian Basin region

2D integrated modelling algorithm was used to calculate the temperature distribution in the lithosphere along the transect IV located in the Western Carpathian-Pannonian Basin area. Based on the determined temperature field and given rheological parameters of the rocks, it was possible to calculate the strength distribution for both compressional and extensional regimes, construct the strength envelopes for chosen columns of the main tectonic units of the model, and thus construct a simple rheological model of the lithosphere along transect IV. The obtained results indicate decrease of the lithospheric strength from the European platform and the Western Carpathians towards the Pannonian Basin. The largest strength (valid for all tectonic units) can be observed within the upper crust with its maxima on the boundary between upper and lower crust, decreasing towards lower crust and disappearing in the lithospheric mantle, suggesting mostly rigid deformation occurring in the upper crust. A local increase in the values of strength can be observed in the eastern segment of the Western Carpathians where crustal thickening accompanies the lithospheric thickening (formation of the lithospheric root), unlike the previous models along transects I and II, that pass through the western segment of the Western Carpathians and their lithosphere-asthenosphere boundary is almost flat and therefore no accompanying crustal thickening is observed and the decrease in strength is slow and steady.

Non-Rigid Multi-Modal 3D Medical Image Registration Based on Foveated Modality Independent Neighborhood Descriptor

The non-rigid multi-modal three-dimensional (3D) medical image registration is highly challenging due to the difficulty in the construction of similarity measure and the solution of non-rigid transformation parameters. A novel structural representation based registration method is proposed to address these problems. Firstly, an improved modality independent neighborhood descriptor (MIND) that is based on the foveated nonlocal self-similarity is designed for the effective structural representations of 3D medical images to transform multi-modal image registration into mono-modal one. The sum of absolute differences between structural representations is computed as the similarity measure. Subsequently, the foveated MIND based spatial constraint is introduced into the Markov random field (MRF) optimization to reduce the number of transformation parameters and restrict the calculation of the energy function in the image region involving non-rigid deformation. Finally, the accurate and efficient 3D medical image registration is realized by minimizing the similarity measure based MRF energy function. Extensive experiments on 3D positron emission tomography (PET), computed tomography (CT), T1, T2, and PD weighted magnetic resonance (MR) images with synthetic deformation demonstrate that the proposed method has higher computational efficiency and registration accuracy in terms of target registration error (TRE) than the registration methods that are based on the hybrid L-BFGS-B and cat swarm optimization (HLCSO), the sum of squared differences on entropy images, the MIND, and the self-similarity context (SSC) descriptor, except that it provides slightly bigger TRE than the HLCSO for CT-PET image registration. Experiments on real MR and ultrasound images with unknown deformation have also be done to demonstrate the practicality and superiority of the proposed method.