Spectral Distribution Aware Image Generation

02/02/2021

Spectral Distribution Aware Image Generation

Steffen Jung, Margret Keuper

Keywords:

Abstract Paper Similar Papers

Abstract: Recent advances in deep generative models for photo-realistic images have led to high quality visual results. Such models learn to generate data from a given training distribution such that generated images can not be easily distinguished from real images by the human eye. Yet, recent work on the detection of such fake images pointed out that they are actually easily distinguishable by artifacts in their frequency spectra. In this paper, we propose to generate images according to the frequency distribution of the real data by employing a spectral discriminator. The proposed discriminator is lightweight, modular and works stably with different commonly used GAN losses. We show that the resulting models can better generate images with realistic frequency spectra, which are thus harder to detect by this cue.

The video of this talk cannot be embedded. You can watch it here:

https://slideslive.com/38948045

(Link will open in new window)

0

0

0

0

Share

This is an embedded video. Talk and the respective paper are published at AAAI 2021 virtual conference. If you are one of the authors of the paper and want to manage your upload, see the question "My papertalk has been externally embedded..." in the FAQ section.

Comments

Post Comment

no comments yet

Similar Papers

06/12/2021

Data-Efficient Instance Generation from Instance Discrimination

Ceyuan Yang, Yujun Shen, Yinghao Xu, Bolei Zhou

Keywords Paper

machine learning, generative model

0

0

0

0

6:53

14/06/2020

Image Processing Using Multi-Code GAN Prior

Jinjin Gu, Yujun Shen, Bolei Zhou

Keywords Paper

generative model, generative adversarial networks, network inversion, gan inversion, deep model prior, image restoration, image super-resolution, image inpainting, image colorization, semantic manipulation

0

0

0

0

1:01

06/12/2021

Deceive D: Adaptive Pseudo Augmentation for GAN Training with Limited Data

Liming Jiang, Bo Dai, Wayne Wu, Chen Change Loy

Keywords Paper

generative model

0

0

0

0

13:51

14/06/2020

PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models

Sachit Menon, Alexandru Damian, Shijia Hu and
Nikhil Ravi, Cynthia Rudin

Keywords Paper

perceptual super-resolution, face hallucination, upsampling, generative models, gan, self-supervision

0

0

0

0

1:01

26/04/2020

Real or Not Real, that is the Question

Yuanbo Xiangli, Yubin Deng, Bo Dai* and
Chen Change Loy, Dahua Lin

Keywords Paper

GAN, generalization, realness, loss function

0

0

0

0

5:07

14/06/2020

Towards Achieving Adversarial Robustness by Enforcing Feature Consistency Across Bit Planes

Sravanti Addepalli, Vivek B.S., Arya Baburaj and
Gaurang Sriramanan, R. Venkatesh Babu

Keywords Paper

adversarial robustness, adversarial defense, adversarial training, fast adversarial training, adversary-free training, adversarial attacks, efficient adversarial training, generalization, feature consistency, deep neural networks

0

0

0

0

1:01

06/12/2020

Differentiable Augmentation for Data-Efficient GAN Training

Shengyu Zhao, Zhijian Liu, Ji Lin and
Jun-Yan Zhu, Song Han

Keywords Paper

0

0

0

0

3:22

17/08/2020

Unsupervised k-modal styled content generation

Omry Sendik, Dani Lischinski, Daniel Cohen-Or

Keywords Paper

StyleGAN, generative adversarial networks, multi-modal distributions

0

0

0

0

11:37

22/11/2021

StyleVideoGAN: A Temporal Generative Model using a Pretrained StyleGAN

Gereon Fox, Ayush Tewari, Mohamed Elgharib, Christian Theobalt

Keywords Paper

video generation, StyleGAN, GAN, embedding, faces, hands, cars, RNN

0

0

0

0

8:07

06/12/2021

Learning to Generate Realistic Noisy Images via Pixel-level Noise-aware Adversarial Training

Yuanhao Cai, Xiaowan Hu, Haoqian Wang and
Yulun Zhang, Hanspeter Pfister, Donglai Wei

Keywords Paper

deep learning, adversarial robustness and security, vision, generative model, graph learning

0

0

0

0

3:05

14/06/2020

Noise Robust Generative Adversarial Networks

Takuhiro Kaneko, Tatsuya Harada

Keywords Paper

generative adversarial networks (gans), image synthesis, noise robust models, image denoising, deep generative models, adversarial training, reparameterization trick, transformation constraint, image restoration, weakly supervised learning

0

0

0

0

1:01

19/08/2021

Beyond the Spectrum: Detecting Deepfakes via Re-Synthesis

Yang He, Ning Yu, Margret Keuper, Mario Fritz

Keywords Paper

Machine Learning, Classification, Deep Learning, Security and Privacy

0

0

0

0

13:25

06/12/2020

Self-Learning Transformations for Improving Gaze and Head Redirection

Yufeng Zheng, Seonwook Park, Xucong Zhang and
Shalini De Mello, Otmar Hilliges

Keywords Paper

0

0

0

0

3:20

06/12/2020

Instance Selection for GANs

Terrance DeVries, Michal Drozdzal, Graham W Taylor

Keywords Paper

0

0

0

0

2:45

03/05/2021

Training GANs with Stronger Augmentations via Contrastive Discriminator

Jongheon Jeong, Jinwoo Shin

Keywords Paper

visual representation learning, contrastive learning, unsupervised learning, data augmentation, generative adversarial networks

0

0

0

0

5:48

26/04/2020

On the "steerability" of generative adversarial networks

Ali Jahanian, Lucy Chai, Phillip Isola

Keywords Paper

generative adversarial network, latent space interpolation, dataset bias, model generalization

0

0

0

0

4:47

06/12/2020

Dual Manifold Adversarial Robustness: Defense against Lp and non-Lp Adversarial Attacks

Wei-An Lin, Chun Pong Lau, Alexander Levine and
Rama Chellappa, Soheil Feizi

Keywords Paper

0

0

0

0

3:24

03/05/2021

You Only Need Adversarial Supervision for Semantic Image Synthesis

Edgar Schoenfeld, Vadim Sushko, Dan Zhang and
Juergen Gall, Bernt Schiele, Anna Khoreva

Keywords Paper

GANs, Semantic Image Synthesis, Image Generation, Deep Learning

0

0

0

0

5:11

05/01/2021

Where to Look?: Mining Complementary Image Regions for Weakly Supervised Object Localization

Sadbhavana Babar, Sukhendu Das

Keywords Paper

0

0

0

0

5:01

05/01/2021

Generating Physically Sound Training Data for Image Recognition of Additively Manufactured Parts

Tobias Nickchen, Stefan Heindorf, Gregor Engels

Keywords Paper

0

0

0

0

4:58

03/05/2021

Robust and Generalizable Visual Representation Learning via Random Convolutions

Zhenlin Xu, Deyi Liu, Junlin Yang and
Colin Raffel, Marc Niethammer

Keywords Paper

robustness, domain generalization, representation learning, data augmentation

0

1

0

0

5:06

14/06/2020

Self-Supervised Monocular Trained Depth Estimation Using Self-Attention and Discrete Disparity Volume

Adrian Johnston, Gustavo Carneiro

Keywords Paper

self-supervised depth estimation, self-supervised learning, self-attention, depth estimation, uncertainty

0

0

0

0

1:01

03/05/2021

Viewmaker Networks: Learning Views for Unsupervised Representation Learning

Alex Tamkin, Mike Wu, Noah Goodman

Keywords Paper

representation learning, self-supervised, views, contrastive learning, unsupervised learning, data augmentation

0

0

0

0

5:03

06/12/2021

Invertible Tabular GANs: Killing Two Birds with One Stone for Tabular Data Synthesis

JAEHOON LEE, Jihyeon Hyeong, Jinsung Jeon and
Noseong Park, Jihoon Cho

Keywords Paper

deep learning, machine learning, robustness, adversarial robustness and security, generative model, privacy

0

0

0

0

10:55

06/12/2021

On the Frequency Bias of Generative Models

Katja Schwarz, Yiyi Liao, Andreas Geiger

Keywords Paper

generative model

0

0

0

0

11:09

12/07/2020

On Leveraging Pretrained GANs for Generation with Limited Data

Miaoyun Zhao, Yulai Cong, Lawrence Carin

Keywords Paper

Deep Learning - General

0

0

0

0

14:22

07/09/2020

E2ETag: An End-to-End Trainable Method for Generating and Detecting Fiducial Markers

John Peace, Eric Psota, Yanfeng Liu, Lance Pérez

Keywords Paper

E2ETag, Fiducial Marker, Pose Estimation, Superimposition, Simulation, Augmentation

0

0

0

0

9:59

14/06/2020

Old Is Gold: Redefining the Adversarially Learned One-Class Classifier Training Paradigm

Muhammad Zaigham Zaheer, Jin-Ha Lee, Marcella Astrid, Seung-Ik Lee

Keywords Paper

anomaly detection, adversarial learning, one-class classification, autoencoder, novelty detection, outlier detection, semi supervised learning, ucsd pedestrian2, mnist, caltech -256

0

0

0

0

1:01

06/12/2021

Encoding Robustness to Image Style via Adversarial Feature Perturbations

Manli Shu, Zuxuan Wu, Micah Goldblum, Tom Goldstein

Keywords Paper

deep learning, machine learning, robustness, adversarial robustness and security, domain adaptation

0

0

0

0

7:36

12/07/2020

Leveraging Frequency Analysis for Deep Fake Image Recognition

Joel Frank, Thorsten Eisenhofer, Lea Schönherr and
Dorothea Kolossa, Thorsten Holz, Asja Fischer

Keywords Paper

Applications - Other

0

0

0

0

12:45

22/11/2021

CamLessMonoDepth: Monocular Depth Estimation with Unknown Camera Parameters

Sai Shyam Chanduri, Igor Vozniak, Zeeshan Khan Suri

Keywords Paper

monocular depth estimation, self-supervised learning, single-camera egomotion, camera intrinsics estimation, sub-pixel convolutions, uncertainty estimation

0

0

0

0

3:02

14/06/2020

Attack to Explain Deep Representation

Mohammad A. A. K. Jalwana, Naveed Akhtar, Mohammed Bennamoun, Ajmal Mian

Keywords Paper

interpreting deep learning, adversarial attack, explanation attack, explainable ai, image generation

0

0

0

0

1:01

16/11/2020

Generation of Realistic Images for Learning in Simulation using FeatureGAN

Nicolas Cruz, Javier Ruiz-del-Solar

Keywords Paper

0

0

0

0

5:02

07/09/2020

Meta-RetinaNet for Few-shot Object Detection

Shaoqi Li, Wenfeng Song, Shuai Li and
Aimin Hao, Hong Qin

Keywords Paper

Few shot, object detection, meta-learning, Meta-RetinaNet, Balanced Loss, coefficient vector

0

0

0

0

8:51

19/08/2021

Enhance Image as You Like with Unpaired Learning

Xiaopeng Sun, Muxingzi Li, Tianyu He, Lubin Fan

Keywords Paper

Computer Vision, 2D and 3D Computer Vision, Applications of Unsupervised Learning

0

0

0

0

11:20

18/07/2021

ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases

Stéphane d'Ascoli, Hugo Touvron, Matthew Leavitt and
Ari Morcos, Giulio Biroli, Levent Sagun

Keywords Paper

Deep Learning, Architectures

0

0

0

0

5:16

19/08/2021

Few-shot Neural Human Performance Rendering from Sparse RGBD Videos

Anqi Pang, Xin Chen, Haimin Luo and
Minye Wu, Jingyi Yu, Lan Xu

Keywords Paper

Computer Vision, 2D and 3D Computer Vision, Biometrics, Face and Gesture Recognition, Motion and Tracking

0

0

0

0

11:02

03/05/2021

Adversarially-Trained Deep Nets Transfer Better: Illustration on Image Classification

Francisco Utrera, Evan Kravitz, N. Benjamin Erichson and
Rajiv Khanna, Michael W Mahoney

Keywords Paper

adversarial training, limited data, influence functions, transfer learning

0

0

0

0

5:12

14/06/2020

Single-Step Adversarial Training With Dropout Scheduling

Vivek B.S., R. Venkatesh Babu

Keywords Paper

adversarial training, robustness, efficient training, representation learning, generalization, supervised learning, recognition, classification, neural networks, deep learning

0

0

0

0

1:01

05/01/2021

G2D: Generate to Detect Anomaly

Masoud Pourreza, Bahram Mohammadi, Mostafa Khaki and
Samir Bouindour, Hichem Snoussi, Mohammad Sabokrou

Keywords Paper

0

0

0

0

5:12