Sökning: "style GAN"

Hittade 2 uppsatser innehållade orden style GAN.

  1. 1. Advanced Data Augmentation : With Generative Adversarial Networks and Computer-Aided Design

    Master-uppsats, Linköpings universitet/Datorseende

    Författare :Ludwig Thaung; [2020]
    Nyckelord :computer vision; machine learning; YOLO; GANs; data augmentation; object recognition; object detection; CAD;

    Sammanfattning : CNN-based (Convolutional Neural Network) visual object detectors often reach human level of accuracy but need to be trained with large amounts of manually annotated data. Collecting and annotating this data can frequently be time-consuming and financially expensive. LÄS MER

  2. 2. Generation of synthetic plant images using deep learning architecture

    Master-uppsats, Blekinge Tekniska Högskola/Institutionen för datavetenskap

    Författare :Ramya Sree Kola; [2019]
    Nyckelord :generative adversarial network; machine learning; style transfer; style GAN; deep learning; FID; Inception score; google vm; leafsnap dataset;

    Sammanfattning : Background: Generative Adversarial Networks (Goodfellow et al., 2014) (GANs)are the current state of the art machine learning data generating systems. Designed with two neural networks in the initial architecture proposal, generator and discriminator. LÄS MER