目录

目录

pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis


<pi-GAN> Pi-gan: Periodic implicit generative adversarial networks for 3d-aware image synthesis

编者按

  • 主要对标、高度对标 GRAF

Motivation

  • StyleGAN类似的noise输入方式(mapping network) + SIREN的周期性激活函数(sinusoidal activation
  • https://longtimenohack.com/posts/paper_reading/2021cvpr_chan_pi/image-20201223163530375.png

Losses

  • discriminator
    • simple ProgressiveGA-like convolutional discriminator;

Overview

FiLM:另外一种input noise使用方式: feature-wise linear modulation

  • 就是首先把 latent 通过mapping变成 $\gamma$ 和 $\beta$ ,然后施加到SIREN的激活函数处
    https://longtimenohack.com/posts/paper_reading/2021cvpr_chan_pi/image-20210311140654298.png
  • 过去的ReLU-based 方法,一般使用concat来condition input noise
  • 作者观察到,对于 SIREN 这种周期性的激活函数来说,condition-by-concatenation 是 sub-optimal的
  • 作者提出,使用 mapping network 进行 feature-wise linear modulation 来 condition SIREN 中的那些Layer
    • [47] arXiv 2017, Film: Visual reasoning with a general conditioning layer.
    • [8] Distlll 2018, Feature-wise transformations. [link]

progressive training

  • 遵循progressiveGAN的方式
  • 先在 低分辨率、大batch size训练,让generator专注于生成 coarse shapes;
  • 然后逐渐增加图像分辨率、给dis添加新层、来辨别fine details
  • 32x32 -> 64x64 -> 128x128
  • 实践中发现,这样的 progressive growing的策略可以在刚开始训练时allow for更大的batch size、allow for higher throuput in images per iteration,对于稳定训练、提速训练有帮助,helped ensure quality and diversity
    • [23] ICLR2018, Progressive growing of GANs for improved quality, stability, and variation.
  • 不需要像progressiveGAN那样增长generator的结构,对于nerf-based生成器,只需要progressively增加采样射线的分辨率即可
  • https://longtimenohack.com/posts/paper_reading/2021cvpr_chan_pi/image-20210310141914872.png

Results

  • https://longtimenohack.com/posts/paper_reading/2021cvpr_chan_pi/image-20201223163713693.png