How to Make a Pizza: Learning a Compositional Layer-Based GAN Model

Papadopoulos, Dim P.; Tamaazousti, Youssef; Ofli, Ferda; Weber, Ingmar; Torralba, Antonio

doi:10.1109/cvpr.2019.00819

Cited by 39 publications

(15 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We conducted an experiment to assist food image generation with pizzaGAN [46], a generative adversarial network (GAN) based model to generate pizza images conditioned by a pizza photo and a cooking instruction (e.g., add corn, or remove ham). All the original images and the manipulated results can be found at http://pizzagan.csail.mit.edu/#.…”

Section: Resultsmentioning

confidence: 99%

Learning to assess visual aesthetics of food images

et al. 2020

View full text Add to dashboard Cite

Distinguishing aesthetically pleasing food photos from others is an important visual analysis task for social media and ranking systems related to food. Nevertheless, aesthetic assessment of food images remains a challenging and relatively unexplored task, largely due to the lack of related food image datasets and practical knowledge. Thus, we present the Gourmet Photography Dataset (GPD), the first large-scale dataset for aesthetic assessment of food photos. It contains 24,000 images with corresponding binary aesthetic labels, covering a large variety of foods and scenes. We also provide a non-stationary regularization method to combat over-fitting and enhance the ability of tuned models to generalize. Quantitative results from extensive experiments, including a generalization ability test, verify that neural networks trained on the GPD achieve comparable performance to human experts on the task of aesthetic assessment. We reveal several valuable findings to support further research and applications related to visual aesthetic analysis of food images. To encourage further research, we have made the GPD publicly available at https://github.com/Openning07/GPA.

show abstract

Section: Resultsmentioning

confidence: 99%

Learning to assess visual aesthetics of food images

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Future work should focus on clustering the samples into consumer groups to strengthen the assumption. Moreover, inspired by Papadopoulos et al (2019), we continue working on rearranging individuals records as layers and generating compositional layer-based synthetic data [27,28]. Another limitation of our study is that we consider the uniqueness of individuals' records as the exposure risk measure.…”

Section: Discussionmentioning

confidence: 99%

Synthesizing Individual Consumers′ Credit Historical Data Using Generative Adversarial Networks

Park

Yoo

2021

Applied Sciences

View full text Add to dashboard Cite

The financial sector accumulates a massive amount of consumer data that contain the most sensitive information daily. These data are strictly limited outside the financial institutions, sometimes even within the same organization, for various reasons such as privacy laws or asset management policy. Financial data has never been more valuable, especially when assessed jointly with data from different industries, including healthcare, insurance, credit bureau, and research institutions. Therefore, it is critical to generate synthetic datasets that retain the statistical or latent properties of the real datasets as well as the privacy protection guaranteed. In this paper, we apply Generative Adversarial Nets (GANs) to generating synthetic consumer credit data to be used for various educational purposes, specifically in developing machine learning models. GAN is preferable to other pseudonymization methods such as masking, swapping, shuffling, or perturbation, for it does not suffer from adding more attributes or data. This study is significant because it is the first attempt to generate the synthetic data of real-world credit data in practical use. The results find that synthetic consumer credit data using GAN shows a substantial utility without severely compromising privacy and would be a useful resource for big data training programs.

show abstract

“…Our Defect-GAN is designed to generate defect samples by simulating the defacement and restoration processes and incorporating randomness to mimic the stochastic variations within defects. Besides, inspired by [64,50,42,69], it deems defects as a special foreground and adopts a layerbased architecture to compose defects on normal samples, thus reserve the normal samples' style and appearance and achieving superior synthesis realism and diversity.…”

Section: Related Workmentioning

confidence: 99%

Defect-GAN: High-Fidelity Defect Synthesis for Automated Defect Inspection

Zhang

Cui

Hung³

et al. 2021

2021 IEEE Winter Conference on Applications of Computer Vision (WACV)

View full text Add to dashboard Cite

Automated defect inspection is critical for effective and efficient maintenance, repair, and operations in advanced manufacturing. On the other hand, automated defect inspection is often constrained by the lack of defect samples, especially when we adopt deep neural networks for this task. This paper presents Defect-GAN, an automated defect synthesis network that generates realistic and diverse defect samples for training accurate and robust defect inspection networks. Defect-GAN learns through defacement and restoration processes, where the defacement generates defects on normal surface images while the restoration removes defects to generate normal images. It employs a novel compositional layer-based architecture for generating realistic defects within various image backgrounds with different textures and appearances. It can also mimic the stochastic variations of defects and offer flexible control over the locations and categories of the generated defects within the image background. Extensive experiments show that Defect-GAN is capable of synthesizing various defects with superior diversity and fidelity. In addition, the synthesized defect samples demonstrate their effectiveness in training better defect inspection networks.

show abstract

How to Make a Pizza: Learning a Compositional Layer-Based GAN Model

Cited by 39 publications

References 51 publications

Learning to assess visual aesthetics of food images

Learning to assess visual aesthetics of food images

Synthesizing Individual Consumers′ Credit Historical Data Using Generative Adversarial Networks

Defect-GAN: High-Fidelity Defect Synthesis for Automated Defect Inspection

Contact Info

Product

Resources

About