Применение нейронных сетей в компьютерной графике [Articol]

Abstract

The article focuses on sketch-based 3D-modeling. While drawings are easy for humans to create, they are very challenging for computers to interpret due to their ambiguity. My approach is to use neural networks to solve image recognition problems. Neural network is able to recognize a sketch and return parameters to generate a polygonal model. The result of practical work is an application written in JavaScript that combines a neural network and a visualizer of three-dimensional models.