User Discrimination of Content Produced by Generative Adversarial Networks

Caporusso, Nicholas; Zhang, Kelei; Carlson, Gordon; Jachetta, Daniel; Patchin, Devon; Romeiser, Spencer; Vaughn, Noah; Walters, Angela

doi:10.1007/978-3-030-25629-6_113

Cited by 8 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our findings stand in contrast to prior research on subjective evaluations of computergenerated artwork, which have largely reported a negative bias towards AI art [8][9][10][11][12] . This work has primarily examined the role of authorship attribution in AI art perception, rather than the aesthetic value of the artworks themselves.…”

Section: Discussioncontrasting

confidence: 99%

“…Moreover, there was a positive correlation between the image-pairs' AI-preference and AI-detection scores, suggesting that the same visual features that made the AI-generated artworks more detectable to participants in Experiment 2 also made those artworks more appealing to participants in Experiment 1 13 . This intriguing pattern underscores the role that explicit bias against artificial creations has likely played in prior investigations [8][9][10][11][12] of the aesthetic appeal of AI-generated artworks: When participants do not know the artworks are computer-generated, they freely prefer them.…”

Section: Discussionmentioning

confidence: 68%

See 1 more Smart Citation

Human Perception of Art in the Age of Artificial Intelligence

van Hees,

Grootswagers,

Quek

et al. 2023

Preprint

View full text Add to dashboard Cite

Recent advancements in Artificial Intelligence (AI) have rendered image-synthesis models capable of producing complex artworks that are nearly indistinguishable from human-made works. Here we present the first quantitative assessment of human perception and preference for art generated by OpenAI's DALL·E 2, a leading AI tool for art creation. Participants were presented with pairs of artworks, one human-made and one AI-generated, in either a preference task or a discrimination task. Results revealed a significant preference for AI-generated artworks. At the same time, a separate group of participants were above-chance at detecting the AI-generated work within each pair, indicating a perceptible distinction between human and artificial creative works. This shift in art preference to favour synthetic creations is poised to revolutionise the way we think about art and its value to human society, prompting reflections on authorship, authenticity, and human creativity in the era of generative AI.

show abstract

Section: Discussioncontrasting

confidence: 99%

Section: Discussionmentioning

confidence: 68%

Human Perception of Art in the Age of Artificial Intelligence

van Hees,

Grootswagers,

Quek

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…One neural net, called the generator, creates new instances of data, whereas the discriminator and analyzes the others for authenticity [5] , [6] . With the use of generative adversarial networks, similar or transformed images are generated with the assorted possibilities towards finding out the most suitable patterns in the analytics and knowledge discovery and thereby the overall scenario becomes highly effectual.…”

Section: Introductionmentioning

confidence: 99%

WITHDRAWN: Advanced machine learning-based analytics on COVID-19 data using generative adversarial networks

Kumar

Harshavardhan²,

Bhukya

et al. 2020

Materials Today: Proceedings

View full text Add to dashboard Cite

The domain of medical diagnosis and predictive analytics is one of the key domains of research with enormous dimensions whereby the diseases of different types can be predicted. Nowadays, there is a huge panic of impact and rapid mutation of the COVID-19 virus impression. The world is getting affected by this virus to a huge extent and there is no vaccine developed so far. India is also having more than 10,000 patients with than 300 deceased. The global human community is having around 20 lacs of Coronavirus patients. The Generative Adversarial Network (GAN) is the contemporary high-performance approach in which the use of advanced neural networks is done for the cavernous analytics of the images and multimedia data. In this research work, the analytics of key points from medical images of the COVID-19 dataset is to be presented using which the diagnosis and predictions can be done for the patients. The GANs are used for the generation, transformation as well as presentation of the dataset and key points using advanced deep learning models which can analyze the patterns in the medical images including X-Ray, CT Scan, and many others. Using such approaches with the integration of GANs, the overall predictive analytics can be made high performance aware as compared to the classical neural networks with multiple layers. In this research manuscript, the inscription of work is projected on the benchmark datasets with the advanced scripting so that the predictive mining and knowledge discovery can be done effectively with more accuracy.

show abstract

“…Creating and implementing an efficient system for the classification of genuine and AIgenerated human faces is the main goal of this research Caporusso et al (2019). The proposed methodology involves a comprehensive preprocessing pipeline consisting of two pivotal stages.…”

Section: Introductionmentioning

confidence: 99%

Optimally Configured Generative Adversarial Networks to Distinguish Real and AI- Generated Human Faces

Chinnaiyan

et al. 2024

Preprint

View full text Add to dashboard Cite

Artificial Intelligence (AI) has come a long way in the last several years, especially in terms of producing human-like faces with deep-fake technology. However, the challenge lies in accurately distinguishing between real and AI-generated human faces. As the applications of such technology continue to expand, the need for robust classification methods becomes crucial to ensure ethical and responsible use. Existing Generative Adversarial Networks (GANs) produce increasingly realistic synthetic faces, making it difficult for traditional methods to differentiate between real and generated faces. This poses potential risks in various domains, including security, identity verification, and misinformation. The primary objective of this research is to design an optimally configured GAN capable of distinguishing between real and generated faces and to develop a robust classifier that accurately classifies human faces as either real or generative. The results showcase the effectiveness of the optimally configured GAN model in achieving high accuracy, reaching 95%, in distinguishing between real and AI-generated faces across state-of-the-art techniques. The research contributes to the ethical deployment of AI technologies, safeguards security applications, strengthens identity verification systems, combats misinformation, and fosters public trust in the era of advanced AI.

show abstract

User Discrimination of Content Produced by Generative Adversarial Networks

Cited by 8 publications

References 4 publications

Human Perception of Art in the Age of Artificial Intelligence

Human Perception of Art in the Age of Artificial Intelligence

WITHDRAWN: Advanced machine learning-based analytics on COVID-19 data using generative adversarial networks

Optimally Configured Generative Adversarial Networks to Distinguish Real and AI- Generated Human Faces

Contact Info

Product

Resources

About