Can We Find Neurons that Cause Unrealistic Images in Deep Generative Networks? (2022-01-17T00:00:00.000000Z)

TL;DR

It is empirically shown that rarely activated neurons are related to the failure results of making diverse objects and inducing artifacts, and a correction method is suggested, called `Sequential Ablation’, to repair the defective part of the generated images without high computational cost and manual efforts.

Abstract

Even though Generative Adversarial Networks (GANs) have shown a remarkable ability to generate high-quality images, GANs do not always guarantee the generation of photorealistic images. Occasionally, they generate images that have defective or unnatural objects, which are referred to as `artifacts'. Research to investigate why these artifacts emerge and how they can be detected and removed has yet to be sufficiently carried out. To analyze this, we first hypothesize that rarely activated neurons and frequently activated neurons have different purposes and responsibilities for the progress of generating images. In this study, by analyzing the statistics and the roles for those neurons, we empirically show that rarely activated neurons are related to the failure results of making diverse objects and inducing artifacts. In addition, we suggest a correction method, called `Sequential Ablation’, to repair the defective part of the generated images without high computational cost and manual efforts.

Authors

Hwanil Choi

1 papers

Wonjoon Chang

1 papers

Jaesik Choi

1 papers

TL;DR

Abstract

Authors

References37 items

Dance In the Wild: Monocular Human Animation with Neural Dynamic Appearance Synthesis

BlendGAN: Implicitly GAN Blending for Arbitrary Stylized Face Generation

A Benchmarking Initiative for Audio-domain Music Generation using the FreeSound Loop Dataset

Alias-Free Generative Adversarial Networks

A Latent Transformer for Disentangled Face Editing in Images and Videos

Exploiting Spatial Dimensions of Latent in GAN for Real-time Image Editing

Automatic Correction of Internal Units in Generative Neural Networks

Taming Transformers for High-Resolution Image Synthesis

SofGAN: A Portrait Image Generator with Dynamic Styling

Attentive Semantic Exploring for Manipulated Face Detection

Analyzing and Improving the Image Quality of StyleGAN

Semantic Hierarchy Emerges in Deep Generative Representations for Scene Synthesis

Improved Precision and Recall Metric for Assessing Generative Models

Visualizing and Understanding Generative Adversarial Networks (Extended Abstract)

A Style-Based Generator Architecture for Generative Adversarial Networks

GAN Dissection: Visualizing and Understanding Generative Adversarial Networks

Large Scale GAN Training for High Fidelity Natural Image Synthesis

Progressive Growing of GANs for Improved Quality, Stability, and Variation

GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium

Deconvolution and Checkerboard Artifacts

Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization

Stability

Mitra

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

and Wen Zheng

and Pierre Hellier

and Björn Ommer

ICASSP 2021 - 2021 IEEE International Conference on Acoustics

and Yi-Hsuan Yang

and Youngjung Uh

pages 12873– 12883

https://github

International Journal of Computer Vision

GENERATIVE ADVERSARIAL NETS

Advances in NeurIPS

Proceedings of ICASSP

In Advances in Neural Information Processing Systems

Field of Study

Venue Information

Name

Type

URL

Alternate Names