{"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/reinforced-evolutionary-neural-architecture","title":"Reinforced Evolutionary Neural Architecture Search","arxiv_id":"1808.00193","date":"2018-08-01","proceeding":null,"authors":["Yukang Chen","Gaofeng Meng","Qian Zhang","Shiming Xiang","Chang Huang","Lisen Mu","Xinggang Wang"],"abstract":"Neural Architecture Search (NAS) is an important yet challenging task in\nnetwork design due to its high computational consumption. To address this\nissue, we propose the Reinforced Evolutionary Neural Architecture Search (RE-\nNAS), which is an evolutionary method with the reinforced mutation for NAS. Our\nmethod integrates reinforced mutation into an evolution algorithm for neural\narchitecture exploration, in which a mutation controller is introduced to learn\nthe effects of slight modifications and make mutation actions. The reinforced\nmutation controller guides the model population to evolve efficiently.\nFurthermore, as child models can inherit parameters from their parents during\nevolution, our method requires very limited computational resources. In\nexperiments, we conduct the proposed search method on CIFAR-10 and obtain a\npowerful network architecture, RENASNet. This architecture achieves a\ncompetitive result on CIFAR-10. The explored network architecture is\ntransferable to ImageNet and achieves a new state-of-the-art accuracy, i.e.,\n75.7% top-1 accuracy with 5.36M parameters on mobile ImageNet. We further test\nits performance on semantic segmentation with DeepLabv3 on the PASCAL VOC.\nRENASNet outperforms MobileNet-v1, MobileNet-v2 and NASNet. It achieves 75.83%\nmIOU without being pre-trained on COCO.","url_abs":"http://arxiv.org/abs/1808.00193v3","url_pdf":"http://arxiv.org/pdf/1808.00193v3.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":"reinforced-evolutionary-neural-architecture","repo_url":"https://github.com/yukang2017/RENAS","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"mxnet","reach":{"status":"ok"}}],"tasks":[{"task_slug":"architecture-search","task_name":"Neural Architecture Search"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"aspp","method_name":"ASPP"},{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"deeplabv3","method_name":"DeepLabv3"},{"method_slug":"dilated-convolution","method_name":"Dilated Convolution"},{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"spatial-pyramid-pooling","method_name":"Spatial Pyramid Pooling"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1808.00193","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}