We present an approach for the reconstruction of textured 3D meshes of human heads from one or few views. Since such few-shot reconstruction is underconstrained, it requires prior knowledge which is hard to impose on traditional 3D reconstruction algorithms. In this work, we rely on the recently introduced 3D representation—neural implicit functions—which, being based on neural networks, allows to naturally learn priors about human heads from data, and is directly convertible to textured mesh. Namely, we extend NeuS, a state-of-the-art neural implicit function formulation, to represent multiple objects of a class (human heads in our case) simultaneously. The underlying neural net architecture is designed to learn the commonalities among these objects and to generalize to unseen ones. Our model is trained on just a hundred smartphone videos and does not require any scanned 3D data. Afterwards, the model can fit novel heads in the few-shot or one-shot modes with good results.
30 April 2024 in Eastern-European Journal of Enterprise Technologies
The paper presents the results of the study of the proposed type of variable profile piles. The proposed type of piles is reinforced concrete driven piles segmented in length. Each subsequent section has a radial displacement along the axis of symmetry relative to the previous section. The positive effect of the performance of the proposed pile is to change the nature of the lateral contact of the pile with the soil, and to increase the drag of the soil. The conducted research is aimed at solving the problem related to the relatively low bearing capacity of traditional square section piles. The research was carried out by the method of model tests of piles on a test bench (tray) at a scale of 1:10. The tests were performed for variant pile types in comparison with a standard square section prismatic pile. The adopted dimensions of the pile model allow the use of this tray without the influence of its boundary conditions on the stress-strain state of the soil. A total of 42 tests were performed, 3 tests for each type of piles compared. Evaluation of pile bearing capacity was performed by static loading of pile models with vertical indentation load until failure. According to the results of the investigations, the resistance values of the compared pile types in the soil were obtained, as well as the dependence of bearing capacity changes on the section dimensions and on the rotation angle. According to the results, the optimal pile solution was selected. The bearing capacity of the proposed optimal pile solution exceeds the bearing capacity of the standard driven pile by 22 %. The results obtained allow us to conclude about the influence of the technological solution of the proposed pile type on its serviceability in soil conditions