![\"\"](\"https://lh3.googleusercontent.com/00FkWyGheo2YWZCI_NbcQPXWyhVu6KBEsrR3bwHdN7DA_P3RjGoosxLUdNcCRAc9CFZ8gKasy1z_Nu79yCWXIhWN-JHDQclNcYkIF5q18_TSh5RB8L4FHBSl_S4b_SK92RBxyP-7\")
Here an off-topic question for my knowledge,
\n\nI'm struggling to find some situation where GAN could be useful to help science or for businesses? \nHave you already had the opportunity to use GAN algorithms in your work? In which situation? Or do you have any idea how these algorithms can be useful? \nThank you
", "messages": [ { "id": "565659", "postDate": "07/01/2019 08:41:17", "content": "Here an off-topic question for my knowledge,
\n\nI'm struggling to find some situation where GAN could be useful to help science or for businesses? \nHave you already had the opportunity to use GAN algorithms in your work? In which situation? Or do you have any idea how these algorithms can be useful? \nThank you
", "rawMarkdown": "Here an off-topic question for my knowledge,\n\nI'm struggling to find some situation where GAN could be useful to help science or for businesses? \nHave you already had the opportunity to use GAN algorithms in your work? In which situation? Or do you have any idea how these algorithms can be useful? \nThank you", "votes": null }, { "id": "565719", "postDate": "07/01/2019 09:57:34", "content": "I think generative methods in general (not only GANs) have many incredible uses.\nFor example, yesterday I read this papers A Survey on GANs for Anomaly Detection:
\n\n\n\n\nAnomaly detection is a significant problem faced in several research areas. Detecting and correctly classifying something unseen as anomalous is a challenging problem that has been tackled in many different manners over the years. \n Generative Adversarial Networks (GANs) and the adversarial training process have been recently employed to face this task yielding remarkable results. In this paper we survey the principal GAN-based anomaly detection methods, highlighting their pros and cons. Our contributions are the empirical validation of the main GAN models for anomaly detection, the increase of the experimental results on different datasets and the public release of a complete Open Source toolbox for Anomaly Detection using GANs.
\n
Arguably, the most famous case (and my favourite) is: Living Portraits
\n\nI'll share here all the papers I see about GANs applied to science, industry ... :)
", "rawMarkdown": "I think generative methods in general (not only GANs) have many incredible uses.\nFor example, yesterday I read this papers [A Survey on GANs for Anomaly Detection](https://arxiv.org/abs/1906.11632):\n> \nAnomaly detection is a significant problem faced in several research areas. Detecting and correctly classifying something unseen as anomalous is a challenging problem that has been tackled in many different manners over the years. \nGenerative Adversarial Networks (GANs) and the adversarial training process have been recently employed to face this task yielding remarkable results. In this paper we survey the principal GAN-based anomaly detection methods, highlighting their pros and cons. Our contributions are the empirical validation of the main GAN models for anomaly detection, the increase of the experimental results on different datasets and the public release of a complete Open Source toolbox for Anomaly Detection using GANs.\n\nArguably, the most famous case (and my favourite) is: **Living Portraits**\n\n\n\nI'll share here all the papers I see about GANs applied to science, industry ... :)", "votes": null }, { "id": "565731", "postDate": "07/01/2019 10:12:09", "content": "Hi,\nGAN can be useful for generating images and text, but also for detecting anomalies. An example of this is anoGAN, which is an unsupervised learning method (a deep convolutional generative adversarial network) capable of identifying anomalous images and image segments, that can serve as imaging biomarker candidates. \nThe paper is here : anoGAN
", "rawMarkdown": "Hi,\nGAN can be useful for generating images and text, but also for detecting anomalies. An example of this is anoGAN, which is an unsupervised learning method (a deep convolutional generative adversarial network) capable of identifying anomalous images and image segments, that can serve as imaging biomarker candidates. \nThe paper is here : [anoGAN](https://arxiv.org/pdf/1703.05921.pdf)", "votes": null }, { "id": "566222", "postDate": "07/01/2019 23:18:49", "content": "GAN can be used for highly imbalaced data to generate new samples for the minority classes and hence is a very good alternative to the classical methods like SMOTE. I have already successfully used GAN for imbalanced financial data.
\n\nIt can also be useful for detecting anomalies as already said , where the minority class is (almost) not present on the training data. This paper is a very good example.
", "rawMarkdown": "GAN can be used for highly imbalaced data to generate new samples for the minority classes and hence is a very good alternative to the classical methods like SMOTE. I have already successfully used GAN for imbalanced financial data. \n\nIt can also be useful for detecting anomalies as already said , where the minority class is (almost) not present on the training data. [This paper](https://arxiv.org/pdf/1803.01798.pdf) is a very good example.", "votes": null }, { "id": "567348", "postDate": "07/03/2019 11:27:24", "content": "There are infinite use of GAN I am mentioning some \n- Creating new animation faces \n- As i guess it can be used also for future face look for person\n- You can apply GAN here in dog image generation \n- Changing resolution of Images \n- Here is the list of infinite number of GAN research papers which ic going to be very helpful to everybody \n- https://github.com/hindupuravinash/the-gan-zoo
", "rawMarkdown": "There are infinite use of GAN I am mentioning some \n- Creating new animation faces \n- As i guess it can be used also for future face look for person\n- You can apply GAN here in dog image generation \n- Changing resolution of Images \n- Here is the list of infinite number of GAN research papers which ic going to be very helpful to everybody \n- https://github.com/hindupuravinash/the-gan-zoo", "votes": null } ], "comments": [ { "id": 565719, "author_name": "jesucristo", "author_url": "", "post_date": "07/01/2019 09:57:34", "content": "I think generative methods in general (not only GANs) have many incredible uses.\nFor example, yesterday I read this papers A Survey on GANs for Anomaly Detection:
\n\n\n\n\nAnomaly detection is a significant problem faced in several research areas. Detecting and correctly classifying something unseen as anomalous is a challenging problem that has been tackled in many different manners over the years. \n Generative Adversarial Networks (GANs) and the adversarial training process have been recently employed to face this task yielding remarkable results. In this paper we survey the principal GAN-based anomaly detection methods, highlighting their pros and cons. Our contributions are the empirical validation of the main GAN models for anomaly detection, the increase of the experimental results on different datasets and the public release of a complete Open Source toolbox for Anomaly Detection using GANs.
\n
Arguably, the most famous case (and my favourite) is: Living Portraits
\n\nI'll share here all the papers I see about GANs applied to science, industry ... :)
", "votes": null, "replies": [] }, { "id": 565731, "author_name": "michelkouassi", "author_url": "", "post_date": "07/01/2019 10:12:09", "content": "Hi,\nGAN can be useful for generating images and text, but also for detecting anomalies. An example of this is anoGAN, which is an unsupervised learning method (a deep convolutional generative adversarial network) capable of identifying anomalous images and image segments, that can serve as imaging biomarker candidates. \nThe paper is here : anoGAN
", "votes": null, "replies": [] }, { "id": 566222, "author_name": "serigne", "author_url": "", "post_date": "07/01/2019 23:18:49", "content": "GAN can be used for highly imbalaced data to generate new samples for the minority classes and hence is a very good alternative to the classical methods like SMOTE. I have already successfully used GAN for imbalanced financial data.
\n\nIt can also be useful for detecting anomalies as already said , where the minority class is (almost) not present on the training data. This paper is a very good example.
", "votes": null, "replies": [] }, { "id": 567348, "author_name": "dataraj", "author_url": "", "post_date": "07/03/2019 11:27:24", "content": "There are infinite use of GAN I am mentioning some \n- Creating new animation faces \n- As i guess it can be used also for future face look for person\n- You can apply GAN here in dog image generation \n- Changing resolution of Images \n- Here is the list of infinite number of GAN research papers which ic going to be very helpful to everybody \n- https://github.com/hindupuravinash/the-gan-zoo
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "565659": "Here an off-topic question for my knowledge,\n\nI'm struggling to find some situation where GAN could be useful to help science or for businesses? \nHave you already had the opportunity to use GAN algorithms in your work? In which situation? Or do you have any idea how these algorithms can be useful? \nThank you", "565719": "I think generative methods in general (not only GANs) have many incredible uses.\nFor example, yesterday I read this papers [A Survey on GANs for Anomaly Detection](https://arxiv.org/abs/1906.11632):\n> \nAnomaly detection is a significant problem faced in several research areas. Detecting and correctly classifying something unseen as anomalous is a challenging problem that has been tackled in many different manners over the years. \nGenerative Adversarial Networks (GANs) and the adversarial training process have been recently employed to face this task yielding remarkable results. In this paper we survey the principal GAN-based anomaly detection methods, highlighting their pros and cons. Our contributions are the empirical validation of the main GAN models for anomaly detection, the increase of the experimental results on different datasets and the public release of a complete Open Source toolbox for Anomaly Detection using GANs.\n\nArguably, the most famous case (and my favourite) is: **Living Portraits**\n\n\n\nI'll share here all the papers I see about GANs applied to science, industry ... :)", "565731": "Hi,\nGAN can be useful for generating images and text, but also for detecting anomalies. An example of this is anoGAN, which is an unsupervised learning method (a deep convolutional generative adversarial network) capable of identifying anomalous images and image segments, that can serve as imaging biomarker candidates. \nThe paper is here : [anoGAN](https://arxiv.org/pdf/1703.05921.pdf)", "566222": "GAN can be used for highly imbalaced data to generate new samples for the minority classes and hence is a very good alternative to the classical methods like SMOTE. I have already successfully used GAN for imbalanced financial data. \n\nIt can also be useful for detecting anomalies as already said , where the minority class is (almost) not present on the training data. [This paper](https://arxiv.org/pdf/1803.01798.pdf) is a very good example.", "567348": "There are infinite use of GAN I am mentioning some \n- Creating new animation faces \n- As i guess it can be used also for future face look for person\n- You can apply GAN here in dog image generation \n- Changing resolution of Images \n- Here is the list of infinite number of GAN research papers which ic going to be very helpful to everybody \n- https://github.com/hindupuravinash/the-gan-zoo" }, "source": "meta" }