[{"data":1,"prerenderedAt":53},["ShallowReactive",2],{"content-query-BO2wC4hx7e":3},{"_path":4,"_dir":5,"_draft":6,"_partial":6,"_locale":7,"title":8,"description":9,"date":10,"cover":11,"type":12,"state":13,"body":14,"_type":47,"_id":48,"_source":49,"_file":50,"_stem":51,"_extension":52},"/activities/zh/3688","zh",false,"","直播 | 一看就会！基于昇思MindSpore框架的模型逆向攻击复现全流程","4月13日（本周日）14点，昇思MindSpore Trusted AI SIG邀你来参加！","2025-04-10","https://obs-mindspore-file.obs.cn-north-4.myhuaweicloud.com/file/2025/04/11/c35e0554616741d4a13fe73d72ce07d1.png","activities",1,{"type":15,"children":16,"toc":44},"root",[17,25,31,36],{"type":18,"tag":19,"props":20,"children":22},"element","h1",{"id":21},"直播-一看就会基于昇思mindspore框架的模型逆向攻击复现全流程",[23],{"type":24,"value":8},"text",{"type":18,"tag":26,"props":27,"children":28},"p",{},[29],{"type":24,"value":30},"4月13日（本周日）14点，昇思MindSpore Trusted AI SIG邀你来参加！本期我们将聚焦模型逆向共计复现与隐私保护，基于《Model Inversion Attacks Against Collaborative Inference》论文，深入探讨模型逆向攻击在协作深度学习系统中的应用与风险，就论文中提出的新技术进行介绍和演示，对rMLE、Inverse-Network和无需查询的影子模型重建方法，并展示这些技术在不同攻击场景下的有效性与通用性。",{"type":18,"tag":26,"props":32,"children":33},{},[34],{"type":24,"value":35},"欢迎各位开发者4月13日下午14点，来B站直播间【MindSpore官方】与我们一同讨论~",{"type":18,"tag":26,"props":37,"children":38},{},[39],{"type":18,"tag":40,"props":41,"children":43},"img",{"alt":7,"src":42},"https://obs-mindspore-file.obs.cn-north-4.myhuaweicloud.com/file/2025/04/11/5672ca0cb8354eabb77f82d3b5890383.png",[],{"title":7,"searchDepth":45,"depth":45,"links":46},4,[],"markdown","content:activities:zh:3688.md","content","activities/zh/3688.md","activities/zh/3688","md",1776506029725]