{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/figr-few-shot-image-generation-with-reptile","title":"FIGR: Few-shot Image Generation with Reptile","arxiv_id":"1901.02199","date":"2019-01-08","proceeding":null,"authors":["Louis Clouâtre","Marc Demers"],"abstract":"Generative Adversarial Networks (GAN) boast impressive capacity to generate\nrealistic images. However, like much of the field of deep learning, they\nrequire an inordinate amount of data to produce results, thereby limiting their\nusefulness in generating novelty. In the same vein, recent advances in\nmeta-learning have opened the door to many few-shot learning applications. In\nthe present work, we propose Few-shot Image Generation using Reptile (FIGR), a\nGAN meta-trained with Reptile. Our model successfully generates novel images on\nboth MNIST and Omniglot with as little as 4 images from an unseen class. We\nfurther contribute FIGR-8, a new dataset for few-shot image generation, which\ncontains 1,548,944 icons categorized in over 18,409 classes. Trained on FIGR-8,\ninitial results show that our model can generalize to more advanced concepts\n(such as \"bird\" and \"knife\") from as few as 8 samples from a previously unseen\nclass of images and as little as 10 training steps through those 8 images. This\nwork demonstrates the potential of training a GAN for few-shot image generation\nand aims to set a new benchmark for future work in the domain.","url_abs":"http://arxiv.org/abs/1901.02199v1","url_pdf":"http://arxiv.org/pdf/1901.02199v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"figr-few-shot-image-generation-with-reptile","repo_url":"https://github.com/OctThe16th/FIGR","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"figr-few-shot-image-generation-with-reptile","repo_url":"https://github.com/marcdemers/FIGR-8","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"figr-few-shot-image-generation-with-reptile","repo_url":"https://github.com/hy-zpg/FIGR","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"meta-learning","task_name":"Meta-Learning"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[{"slug":"figr-8","name":"FIGR-8","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1901.02199","atlas_url":"https://app.syntology.ai/?focus=1901.02199","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}