FIELD: computing technology.
SUBSTANCE: technical result is achieved by generating images with a resolution by means of conditionally independent pixel synthesis, wherein a random vector z ∈ Z and the coordinates of pixels (x, y) ∈ {0 ... W-1}×{0 ... H-1} of a floating-point image are extracted by a multilayer perceptron, processing the coordinates of pixels of the floating-point image with a set of sines with different frequencies, searching for pre-trained coordinate embeddings in the network weights using the coordinates of pixels of the floating-point image, transforming a random vector by a display network using the coordinates of pixels of the floating-point image, the output whereof affects the weights in the fully connected layer, returning the RGB values of each pixel of the image by a multilayer perceptron, estimating the perceptron generator G(x; y; z) on each pair (x, y), calculating the full output image according to the estimate I = {G(x, y; z)|(x, y) ∈ mgrid (H, W)}.
EFFECT: improvement in image quality with limited resources of the graphics processors.
6 cl, 3 tbl, 13 dwg
Title | Year | Author | Number |
---|---|---|---|
MODELLING PEOPLE'S CLOTHING BASED ON MULTIPLE POINTS | 2021 |
|
RU2776825C1 |
RAPID TWO-LAYER NEURAL NETWORK SYNTHESIS OF REALISTIC IMAGES OF A NEURAL AVATAR BASED ON A SINGLE IMAGE | 2020 |
|
RU2764144C1 |
METHOD AND COMPUTING DEVICE FOR FORMATION OF A TRUE DISPLAY OF TIME FLOW OF A DIURNAL SCALE | 2020 |
|
RU2745209C1 |
REPEATED SYNTHESIS OF IMAGE USING DIRECT DEFORMATION OF IMAGE, PASS DISCRIMINATOR AND COORDINATE-BASED REMODELLING | 2019 |
|
RU2726160C1 |
NEURAL NETWORK TRANSFER OF THE FACIAL EXPRESSION AND POSITION OF THE HEAD USING HIDDEN POSITION DESCRIPTORS | 2020 |
|
RU2755396C1 |
METHOD FOR CONSTRUCTING A SCENE REPRESENTATION WITH DIRECT CORRECTION FOR REAL-TIME IMAGE SYNTHESIS | 2022 |
|
RU2799237C1 |
METHOD OF SYNTHESIS OF A TWO-DIMENSIONAL IMAGE OF A SCENE VIEWED FROM A REQUIRED VIEW POINT AND ELECTRONIC COMPUTING APPARATUS FOR IMPLEMENTATION THEREOF | 2020 |
|
RU2749749C1 |
METHOD OF CREATING FULL-LENGTH ANIMATED AVATAR OF PERSON FROM ONE IMAGE OF PERSON, COMPUTING DEVICE AND MACHINE-READABLE MEDIUM FOR IMPLEMENTATION THEREOF | 2023 |
|
RU2813485C1 |
METHOD FOR LOCAL GENERATION AND REPRESENTATION OF WALLPAPER STREAM AND COMPUTER IMPLEMENTING IT | 2020 |
|
RU2768551C1 |
TRAINING GAN (GENERATIVE ADVERSARIAL NETWORKS) TO CREATE PIXEL-BY-PIXEL ANNOTATION | 2019 |
|
RU2735148C1 |
Authors
Dates
2022-04-14—Published
2021-03-04—Filed