{"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/saliency-aware-neural-architecture-search","title":"Saliency-Aware Neural Architecture Search","arxiv_id":null,"date":"2022-11-01","proceeding":"NIPS 2022 11","authors":["Ramtin Hosseini","Pengtao Xie"],"abstract":"Recently a wide variety of NAS methods have been proposed and achieved con\u0002siderable success in automatically identifying highly-performing architectures of\r\nneural networks for the sake of reducing the reliance on human experts. Existing\r\nNAS methods ignore the fact that different input data elements (e.g., image pixels)\r\nhave different importance (or saliency) in determining the prediction outcome.\r\nThey treat all data elements as being equally important and therefore lead to subop\u0002timal performance. To address this problem, we propose an end-to-end framework\r\nwhich dynamically detects saliency of input data, reweights data using saliency\r\nmaps, and searches architectures on saliency-reweighted data. Our framework is\r\nbased on four-level optimization, which performs four learning stages in a unified\r\nway. At the first stage, a model is trained with its architecture tentatively fixed. At\r\nthe second stage, saliency maps are generated using the trained model. At the third\r\nstage, the model is retrained on saliency-reweighted data. At the fourth stage, the\r\nmodel is evaluated on a validation set and the architecture is updated by minimizing\r\nthe validation loss. Experiments on several datasets demonstrate the effectiveness\r\nof our framework","url_abs":"https://openreview.net/forum?id=Ho6oWAslz5L","url_pdf":"https://openreview.net/pdf?id=Ho6oWAslz5L","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":"saliency-aware-neural-architecture-search","repo_url":"https://github.com/leopard-ai/betty","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"architecture-search","task_name":"Neural Architecture Search"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}