Generative Adversarial Networks (GANs) were first reported on in 2014 from Ian Goodfellow and others in Yoshua Bengio's lab. [1] Since then, GANs have exploded in popularity. Here are a few examples to check out: Pix2Pix [2], CycleGAN [3], A whole list [4]
\n\nThe idea behind GANs is that you have two networks, a generator and a discriminator, competing against each other. The generator makes fake data to pass to the discriminator. The discriminator also sees real data and predicts if the data it's received is real or fake.
\n\nThe generator is trained to fool the discriminator, it wants to output data that looks as close as possible to real data. And the discriminator is trained to figure out which data is real and which is fake. What ends up happening is that the generator learns to make data that is indistiguishable from real data to the discriminator.
\n\n[1] https://arxiv.org/pdf/1406.2661.pdf\n[2] https://affinelayer.com/pixsrv/\n[3] https://github.com/junyanz/CycleGAN\n[4] https://github.com/wiseodd/generative-models
", "messages": [ { "id": "568031", "postDate": "07/04/2019 10:12:50", "content": "Generative Adversarial Networks (GANs) were first reported on in 2014 from Ian Goodfellow and others in Yoshua Bengio's lab. [1] Since then, GANs have exploded in popularity. Here are a few examples to check out: Pix2Pix [2], CycleGAN [3], A whole list [4]
\n\nThe idea behind GANs is that you have two networks, a generator and a discriminator, competing against each other. The generator makes fake data to pass to the discriminator. The discriminator also sees real data and predicts if the data it's received is real or fake.
\n\nThe generator is trained to fool the discriminator, it wants to output data that looks as close as possible to real data. And the discriminator is trained to figure out which data is real and which is fake. What ends up happening is that the generator learns to make data that is indistiguishable from real data to the discriminator.
\n\n[1] https://arxiv.org/pdf/1406.2661.pdf\n[2] https://affinelayer.com/pixsrv/\n[3] https://github.com/junyanz/CycleGAN\n[4] https://github.com/wiseodd/generative-models
", "rawMarkdown": "Generative Adversarial Networks (GANs) were first reported on in 2014 from Ian Goodfellow and others in Yoshua Bengio's lab. [1] Since then, GANs have exploded in popularity. Here are a few examples to check out: Pix2Pix [2], CycleGAN [3], A whole list [4]\n\nThe idea behind GANs is that you have two networks, a generator and a discriminator, competing against each other. The generator makes fake data to pass to the discriminator. The discriminator also sees real data and predicts if the data it's received is real or fake. \n\nThe generator is trained to fool the discriminator, it wants to output data that looks as close as possible to real data. And the discriminator is trained to figure out which data is real and which is fake. What ends up happening is that the generator learns to make data that is indistiguishable from real data to the discriminator.\n\n[1] https://arxiv.org/pdf/1406.2661.pdf\n[2] https://affinelayer.com/pixsrv/\n[3] https://github.com/junyanz/CycleGAN\n[4] https://github.com/wiseodd/generative-models", "votes": null }, { "id": "568079", "postDate": "07/04/2019 11:16:37", "content": "thanks for sharing 👍
", "rawMarkdown": "thanks for sharing 👍", "votes": null }, { "id": "568142", "postDate": "07/04/2019 13:23:10", "content": "Thanks @grapestone5321 for sharing but it would be really helpful if share a detailed and conscience explanation of GAN as all the research papers makes use of advanced mathematics but it get really complicated to replicate them in codes. (Eg. Earth movers distances used in WGAN)
", "rawMarkdown": "Thanks @grapestone5321 for sharing but it would be really helpful if share a detailed and conscience explanation of GAN as all the research papers makes use of advanced mathematics but it get really complicated to replicate them in codes. (Eg. Earth movers distances used in WGAN)", "votes": null }, { "id": "568636", "postDate": "07/05/2019 08:49:25", "content": "Great summary and links, thanks!
", "rawMarkdown": "Great summary and links, thanks!", "votes": null }, { "id": "570560", "postDate": "07/08/2019 13:08:29", "content": "Refer to the kernel: https://www.kaggle.com/grapestone5321/gan-basic-using-mnist
", "rawMarkdown": "Refer to the kernel: https://www.kaggle.com/grapestone5321/gan-basic-using-mnist", "votes": null }, { "id": "572021", "postDate": "07/10/2019 11:03:13", "content": "Also refer to the DCGAN Basic kernel: https://www.kaggle.com/grapestone5321/dcgan-basic-using-svhn
", "rawMarkdown": "Also refer to the DCGAN Basic kernel: https://www.kaggle.com/grapestone5321/dcgan-basic-using-svhn", "votes": null } ], "comments": [ { "id": 568079, "author_name": "iluvmahheart", "author_url": "", "post_date": "07/04/2019 11:16:37", "content": "thanks for sharing 👍
", "votes": null, "replies": [] }, { "id": 568142, "author_name": "thanatoz", "author_url": "", "post_date": "07/04/2019 13:23:10", "content": "Thanks @grapestone5321 for sharing but it would be really helpful if share a detailed and conscience explanation of GAN as all the research papers makes use of advanced mathematics but it get really complicated to replicate them in codes. (Eg. Earth movers distances used in WGAN)
", "votes": null, "replies": [] }, { "id": 568636, "author_name": "edtallison", "author_url": "", "post_date": "07/05/2019 08:49:25", "content": "Great summary and links, thanks!
", "votes": null, "replies": [] }, { "id": 570560, "author_name": "grapestone5321", "author_url": "", "post_date": "07/08/2019 13:08:29", "content": "Refer to the kernel: https://www.kaggle.com/grapestone5321/gan-basic-using-mnist
", "votes": null, "replies": [] }, { "id": 572021, "author_name": "grapestone5321", "author_url": "", "post_date": "07/10/2019 11:03:13", "content": "Also refer to the DCGAN Basic kernel: https://www.kaggle.com/grapestone5321/dcgan-basic-using-svhn
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "568031": "Generative Adversarial Networks (GANs) were first reported on in 2014 from Ian Goodfellow and others in Yoshua Bengio's lab. [1] Since then, GANs have exploded in popularity. Here are a few examples to check out: Pix2Pix [2], CycleGAN [3], A whole list [4]\n\nThe idea behind GANs is that you have two networks, a generator and a discriminator, competing against each other. The generator makes fake data to pass to the discriminator. The discriminator also sees real data and predicts if the data it's received is real or fake. \n\nThe generator is trained to fool the discriminator, it wants to output data that looks as close as possible to real data. And the discriminator is trained to figure out which data is real and which is fake. What ends up happening is that the generator learns to make data that is indistiguishable from real data to the discriminator.\n\n[1] https://arxiv.org/pdf/1406.2661.pdf\n[2] https://affinelayer.com/pixsrv/\n[3] https://github.com/junyanz/CycleGAN\n[4] https://github.com/wiseodd/generative-models", "568079": "thanks for sharing 👍", "568142": "Thanks @grapestone5321 for sharing but it would be really helpful if share a detailed and conscience explanation of GAN as all the research papers makes use of advanced mathematics but it get really complicated to replicate them in codes. (Eg. Earth movers distances used in WGAN)", "568636": "Great summary and links, thanks!", "570560": "Refer to the kernel: https://www.kaggle.com/grapestone5321/gan-basic-using-mnist", "572021": "Also refer to the DCGAN Basic kernel: https://www.kaggle.com/grapestone5321/dcgan-basic-using-svhn" }, "source": "meta" }