1,227 Hits in 4.7 sec

Synthesizing Visual Illusions Using Generative Adversarial Networks [article]

Alexander Gomez-Villa, Adrian Martín, Javier Vazquez-Corral, Jesús Malo, Marcelo Bertalmío
2019 arXiv   pre-print
It takes the form of a generative adversarial network, with a generator of visual illusion candidates and two discriminator modules, one for the inducer background and another that decides whether or not  ...  In this work we introduce the first ever framework to generate novel visual illusions with an artificial neural network (ANN).  ...  We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research.  ... 
arXiv:1911.09599v1 fatcat:bsxyisecs5fnjlww2zlfe3tuwq

Is current research on adversarial robustness addressing the right problem? [article]

Ali Borji
2022 arXiv   pre-print
Indeed, research on adversarial robustness has led to invaluable insights helping us understand and explore different aspects of the problem.  ...  Maybe instead of narrowing down on imperceptible adversarial perturbations, we should attack a more general problem which is finding architectures that are simultaneously robust to perceptible perturbations  ...  The nature of visual illusions is very different from adversarial examples. Our visual system has evolved to enable us to function with high reliability without making deadly mistakes.  ... 
arXiv:2208.00539v2 fatcat:m5mwg4pdpnfjxaajosw6yovj7m

DeepFake Detection: Current Challenges and Next Steps [article]

Siwei Lyu
2020 arXiv   pre-print
High quality fake videos and audios generated by AI-algorithms (the deep fakes) have started to challenge the status of videos and audios as definitive evidence of events.  ...  Our aim is to use the adversarial perturbations (amplified by 30 for better visualization) to distract DNN-based face detectors, such that the quality of the obtained face set as training data to the AI  ...  The data-driven deep neural network based DeepFake detection methods are particularly susceptible to anti-forensic attacks due to the known vulnerability of general deep neural network classification models  ... 
arXiv:2003.09234v1 fatcat:nv2sq3mfl5dzpj3bvfrmi6biaa

Learnable Visual Markers [article]

Oleg Grinchuk, Vadim Lebedev, Victor Lempitsky
2016 arXiv   pre-print
We propose a new approach to designing visual markers (analogous to QR-codes, markers for augmented reality, and robotic fiducial tags) based on the advances in deep generative networks.  ...  In our approach, the markers are obtained as color images synthesized by a deep network from input bit strings, whereas another deep network is trained to recover the bit strings back from the photos of  ...  Towards this end [27] generate examples that maximize probabilities of certain classes according to the network, [33] generate visual illusions that maximize such probabilities while retaining similarity  ... 
arXiv:1610.09237v1 fatcat:6ftk6g5jxjbfdpqqv2hrhebm5q

Toward the Creation and Obstruction of DeepFakes [chapter]

Yuezun Li, Pu Sun, Honggang Qi, Siwei Lyu
2022 Advances in Computer Vision and Pattern Recognition  
We present a new large-scale challenging DeepFake video dataset, Celeb-DF, which contains 5, 639 high-quality DeepFake videos of celebrities generated using an improved synthesis process.  ...  However, current DeepFake datasets suffer from low visual quality and do not resemble DeepFake videos circulated on the Internet.  ...  ), we also test the black-box attack using the adversarial perturbation generated from one landmark extractor to attack other extractors.  ... 
doi:10.1007/978-3-030-87664-7_4 fatcat:6yuopmliazevfcdps6jw5enhky

InGAN: Capturing and Retargeting the "DNA" of a Natural Image

Assaf Shocher, Shai Bagon, Phillip Isola, Michal Irani
2019 2019 IEEE/CVF International Conference on Computer Vision (ICCV)  
Generative Adversarial Networks (GANs) typically learn a distribution of images in a large image dataset, and are then able to generate new images from this distribution.  ...  It is then able to synthesize a plethora of new natural images of significantly different sizes, shapes and aspect-ratios all with the same internal patch-distribution (same "DNA") as the input image.  ...  [12, 11] used pretrained network features to synthesize textures.  ... 
doi:10.1109/iccv.2019.00459 dblp:conf/iccv/ShocherBII19 fatcat:jlj5vsmttzgkjabzsqjclsou2u

Generative Digital Humanities

Fabian Offert, Peter Bell
2020 Workshop on Computational Humanities Research  
In this paper, we examine generative adversarial networks, a state-of-the art generative machine learning technique.  ...  If "all models are wrong, some are useful", as the often-cited passage reads, we argue that, in case of the digital humanities, the most useful-wrong models are generative.  ...  The original paper by Goodfellow demonstrates the potential of generative adversarial networks to synthesize images in particular by synthesizing new handwritten digits from the MNIST dataset.  ... 
dblp:conf/chr/Offert020 fatcat:szivogi7jnevdhpb22gkcudjla

Pixel Sampling for Style Preserving Face Pose Editing [article]

Xiangnan Yin, Di Huang, Hongyu Yang, Zehua Fu, Yunhong Wang, Liming Chen
2021 arXiv   pre-print
By leveraging high-dimensional embedding at the inpainting stage, finer details are generated.  ...  In this paper, we take advantage of the well-known frontal/profile optical illusion and present a novel two-stage approach to solve the aforementioned dilemma, where the task of face pose manipulation  ...  Related Work Generative Adversarial Network (GAN) In recent years, Generative Adversarial Networks (GAN) has been one of the most popular research directions for image generation.  ... 
arXiv:2106.07310v1 fatcat:xo3dn2ldzfd6pbkrtkxreuolsu

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

Eric R. Chan, Marco Monteiro, Petr Kellnhofer, Jiajun Wu, Gordon Wetzstein
2021 arXiv   pre-print
We propose a novel generative model, named Periodic Implicit Generative Adversarial Networks (π-GAN or pi-GAN), for high-quality 3D-aware image synthesis. π-GAN leverages neural representations with periodic  ...  We have witnessed rapid progress on 3D-aware image synthesis, leveraging recent advances in generative visual models and neural rendering.  ...  Visual Object Networks [72] and PrGANs [11] learn to synthesize 2D images by first generating a voxelized 3D shape using a 3D-GAN [67] and then projecting it into 2D.  ... 
arXiv:2012.00926v2 fatcat:scgeqv3a4ngh7n4qzu2tkst5aq

Visual Indeterminacy in GAN Art

Aaron Hertzmann
2020 Leonardo: Journal of the International Society for the Arts, Sciences and Technology  
This paper explores visual indeterminacy as a description for artwork created with Generative Adversarial Networks (GANs).  ...  Visual indeterminacy describes images that appear to depict real scenes, but on closer examination, defy coherent spatial interpretation.  ...  The first widespread trend in machine learning-based artwork is the use of Generative Adversarial Networks (GANs) [1, 2] .  ... 
doi:10.1162/leon_a_01930 fatcat:sbvpsvm67bbtpdhbshiuaajeum

InGAN: Capturing and Remapping the "DNA" of a Natural Image [article]

Assaf Shocher, Shai Bagon, Phillip Isola, Michal Irani
2019 arXiv   pre-print
Generative Adversarial Networks (GANs) typically learn a distribution of images in a large image dataset, and are then able to generate new images from this distribution.  ...  It is then able to synthesize a plethora of new natural images of significantly different sizes, shapes and aspect-ratios - all with the same internal patch-distribution (same "DNA") as the input image  ...  [12, 11] used pretrained network features to synthesize textures.  ... 
arXiv:1812.00231v2 fatcat:sfiusmquzjd3rgs45lryz4l24u

GRAM: Generative Radiance Manifolds for 3D-Aware Image Generation [article]

Yu Deng, Jiaolong Yang, Jianfeng Xiang, Xin Tong
2021 arXiv   pre-print
For each viewing ray, we calculate ray-surface intersections and accumulate their radiance generated by the network.  ...  By training and rendering such radiance manifolds, our generator can produce high quality images with realistic fine details and strong visual 3D consistency.  ...  Introduction Learning 3D-aware image generation with Generative Adversarial Networks (GAN) [20] has attracted a surge of attention in recent years [11, 13, 15, 24, 34, [44] [45] [46] 57] .  ... 
arXiv:2112.08867v2 fatcat:jp2pnqzkojdbziuvztpgj4gwcy

Enhancing Quality of Pose-varied Face Restoration with Local Weak Feature Sensing and GAN Prior [article]

Kai Hu, Yu Liu, Renhe Liu, Wei Lu, Gang Yu, Bin Fu
2022 arXiv   pre-print
Facial semantic guidance (including facial landmarks, facial heatmaps, and facial parsing maps) and facial generative adversarial networks (GAN) prior have been widely used in blind face restoration (BFR  ...  In this work, we propose a well-designed blind face restoration network with generative facial prior. The proposed network is mainly comprised of an asymmetric codec and a StyleGAN2 prior network.  ...  For each region, we build upon the same discriminator network and employ different local losses for adversarial training.  ... 
arXiv:2205.14377v2 fatcat:tztygp77vjevfi75r4542cfuce

MU-GAN: Facial Attribute Editing based on Multi-attention Mechanism [article]

Ke Zhang, Yukun Su, Xiwang Guo, Liang Qi, Zhenbing Zhao
2020 arXiv   pre-print
In this work, we propose a Multi-attention U-Net-based Generative Adversarial Network (MU-GAN).  ...  First, we replace a classic convolutional encoder-decoder with a symmetric U-Net-like structure in a generator, and then apply an additive attention mechanism to build attention-based U-Net connections  ...  ACKNOWLEDGMENT The authors gratefully acknowledge the support of NVIDIA Corporation with the donation of the GPU used for this research.  ... 
arXiv:2009.04177v1 fatcat:p3xdfb5w3zetfm2cvtx52h5enq

fMRI Brain Decoding and Its Applications in Brain–Computer Interface: A Survey

Bing Du, Xiaomu Cheng, Yiping Duan, Huansheng Ning
2022 Brain Sciences  
With the great success of deep learning on image recognition and generation, deep neural networks (DNN) have been engaged in reconstructing visual stimuli from human brain activity via functional magnetic  ...  Specifically, we focused on current brain activity decoding models with high attention: variational auto-encoder (VAE), generative confrontation network (GAN), and the graph convolutional network (GCN)  ...  GAN-Based Brain Decoding GAN consists of a generative network and a discriminant network, and tends to generate high-quality pictures through adversarial training of the discriminator [81] .  ... 
doi:10.3390/brainsci12020228 pmid:35203991 pmcid:PMC8869956 fatcat:t664eccq6nh5plnvhac2r2gcpa
« Previous Showing results 1 — 15 out of 1,227 results